Chapter 5: Barriers to the Growth of Blended Learning
If existing blended-learning schools are getting good results, why hasn’t the practice spread more widely? One obvious answer is that the current schools are very new. Hardly any have more than just a few years of experience. Moreover, starting a new school—or radically transforming an older school’s structure—is a complicated undertaking that is not for the faint of heart.
Merit Prep in Newark opened its doors in August 2012 as one of New Jersey’s first two schools blending in-person and online instruction. Founder Ben Rayer has his work cut out for him: While a sign on the green-accented walls of this school across from City Hall says, “To mastery and beyond,” the beginning reality is that just 17 percent of his 80 new sixth-grade students are performing at grade level.
The office building the school is occupying already had a wide central space (Merit Prep teachers and students call it “the stadium”), so there were few architectural difficulties in getting the facility up and running around a computer-filled core. A more serious existential challenge threatened this new charter school, though: a suit from the state teachers union, known as the New Jersey Education Association, asking for an injunction that would prevent the school even from opening. Their argument was that New Jersey’s Department of Education lacked the authority to authorize online charter schools—even though Merit Prep’s students are most emphatically in a school building, and learning from flesh and blood teachers, a scenario that has little in common with the virtual schools operating in several other states.
A judge denied the union’s motion, but did so without ruling on the merits of the suit, which will grind forward over the next year, a cloud over this burst of idealism. Needless to say, the possibility of facing a major lawsuit from opponents of new methods can dampen enthusiasm for starting new styles of schools. And that’s not the only obstacle standing in the way of more blended-learning experiments. Here are a few other major reasons why blended learning is not yet available on a large scale.
Bottleneck 1: A Lack of Research
While blended-learning proponents can point to some initially good test scores, there is little solid data published in peer-reviewed journals. “Honestly, it’s so early on, no one knows what works and doesn’t work,” says Diane Tavenner, leader of Summit Public Schools. “Indeed,” notes Scott Benson, who directs blended-learning grants at the Bill & Melinda Gates Foundation, “part of me is really nervous—that the dialogue and enthusiasm is outpacing the results.”
Education involves humans, and human outcomes are affected by so many variables that it’s hard to pin any result to one thing. The fact that one teacher got good test results in a blended classroom may mean that blended learning worked. Or it may mean that you had a great teacher who was willing to try new things. It sounds impressive to know that at KIPP Empower Academy, the kindergartners went from 64 percent basic or below basic achievement on the STEP literacy assessment to just 4 percent basic or below basic, and 96 percent proficient or advanced. It also sounds good that in the Khan Academy Los Altos pilot, twice as many seventh-grade students in developmental math were at grade level at the end of the year as at the beginning. Yet you can’t assert that any other school that tries the same thing will get the same results.
One ongoing project in blended learning is to gather more good data. Cheryl Niehaus, an education program officer at the Michael & Susan Dell Foundation who follows blended learning closely, says, “Last year, one thing stood out: how much activity there was, how much excitement. And also how relatively little information there is on what works for students in a blended-learning model.” To help address this, Dell commissioned its five white papers on blended learning schools (please see chapter 3) and a study from the Stanford Research Institute on quantitative results from those schools.
There are a few pieces of sophisticated evidence. One meta-analysis of blended-learning studies looked at research published between 1996 and 2008 that met minimum quality standards. The results were mixed, but 11 of the 51 studies favored online or blended learning, while two favored face-to-face education.
Brian Greenberg who was the chief academic officer of Envision Schools before joining the Silicon Schools Fund, ran a study comparing blended learning and traditional instruction in a five-week summer-school program in Oakland for students who had failed algebra. The same teacher taught both sections, and children were randomly assigned to the blended-learning class or a traditional class. At the end of the program, the children from the blended-learning section showed slightly more improvement from their pre-test to their post-test. “The gains themselves were not particularly robust,” says Greenberg, but what was interesting is that the test was a “pure algebra measure.” Many of the kids in the blended learning section “spent a good portion of their summer on pre-algebra skills,” as the data quickly revealed that they had all sorts of gaps in their pre-algebra math knowledge, which may have been why they failed algebra in the first place. The value of filling in those gaps “doesn’t come up if you’re only testing algebra,” but it should show itself in other math work by those students. Greenberg’s summary is that the blended kids “spent less time on algebra than the regular kids, yet did as well or better” on an algebra assessment. Based on that result, he’s since been working on rolling out broader blended-learning programs, most notably in all the schools launched by the Silicon Schools Fund, a fund financed by the Fisher Fund and other donors to start blended schools in the San Francisco Bay area.
For its remedial math class, Arizona State University in 2011 launched a pilot using Knewton, an adaptive learning software platform. According to Knewton COO David Liu, withdrawal rates dropped by 56 percent, pass rates increased, and 45 percent of the students finished the class early. That last statistic hadn’t been kept before, because it wasn’t possible until the daily assessment and instant dashboard postings of the blended model made it possible to know when individual students had reached mastery. Given that long stints in remedial classes can lessen the chance that a student finishes college, anything that speeds the process along could have outsized results.
In the long run, this ability of instructors to see specific things in individual student performance that were invisible before will be one of blended learning’s greatest assets. A few weeks before California’s statewide assessments, Juan Nuñez at Thurgood Marshall learned, thanks to the constant streams of data he was receiving from his computer programs, that his students were struggling with two skills that were going to be tested. Looking back, he could see that these skills hadn’t been covered adequately in the curriculum. So he decided, “let me take this week to go ahead and teach that explicitly.” Without the feedback that his students didn’t know those skills, he wouldn’t have gone back to fill that gap, and his students’ understanding and performance outcomes would have been lower.
Even if there isn’t yet much official overarching data on what works, the first bit of good news is that multiple new studies are in process right now, with many of the results due in 2013. Financing research can be a philanthropic sweet spot since it is an area that people busy running schools tend to underinvest in. The field of education spends shockingly small portions of overall budget on R&D—by some measures, less than 1 percent (compared to double digits in the pharmaceutical industry, for instance). One result of that discrepancy is that while a doctor facing a common condition—perhaps childhood asthma—is starting to have a well-tested protocol of treatment to apply, educators facing similarly common conditions do not. Philanthropists can help put funds where education policymakers have failed to.
A final bit of good news is that blended learning is itself founded on data and measurement, so its practitioners are able to make good decisions about efficacy long before they see the results of complex multi-year double-blind studies. Over time, the statistical measures will get richer. And given the initial results, blended-learning proponents are bullish.
Bottleneck 2: The Dilbert Reaction
Another bottleneck slowing blended learning has to do with what it looks like compared to the archetypal vision of education, or at least education as grown-ups wax nostalgic about it. “There’s a lot of ‘small c’ conservatism” among parents and educators, says Michael Horn of the Innosight Institute. They say, “This is how I went to school. Why shouldn’t my kids have the same experience?”
Even people who like the idea of technology in schools can have a visceral reaction to pictures of children sitting in what looks like a cubicle farm in Scott Adams’ Dilbert comic strip. Grown-ups don’t like logging time in cubicles, so why would they want that for their children? Even some tech titans resist that for their children. An October 2011 story in the New York Times described a Waldorf School in Silicon Valley—which attracts offspring of people like the chief technology officer of eBay—where screens are not allowed in the classroom, and educators believe that “computers inhibit creative thinking, movement, human interaction, and attention spans.”
We all have romantic notions of what education looks like, often involving wooden desks, dusty books, a chalkboard, and a teacher in a houndstooth blazer. Blended learning doesn’t look like that. High pupil-to-teacher ratios are also jarring for parents who have had the idea drummed into them that smaller classes must be better. “No laptop can replace a teacher” is a common refrain that captures this visceral reaction.
While it is certainly true that computers and teachers are far from interchangeable, we live in an era that increasingly differentiates between the core aspects of a job and the peripheral elements. Robots weld car frames, while humans adjust door fits. Computers collect basic information, then customer-service reps swoop in to answer questions on what insurance policy is right for you.
A laptop most definitely can replace a teacher when it comes to taking attendance, grading quizzes, conducting performance assessments, and presenting basic content. If a teacher doesn’t have to spend time on these things, she can focus, instead, on what might actually be her core capabilities: helping children learn how to learn, and inspiring them to keep at it. It is this extra teacher time and attention that proponents of blended learning find parents most need to see to believe. Providing visual evidence of these schools working is one reason the Jaquelin Hume Foundation funded the creation of videos of several blended-learning institutions, and brings journalists, legislators, and community leaders on site visits to see the classrooms for themselves. CityBridge Foundation’s Education Innovation Fellowship likewise brings teachers to blended classrooms so they can observe this individualized attention first hand. When parents do see it—even parents who aren’t sending their kids to failing schools—many are quite intrigued.
The evolution of computers into more intimate and responsive forms may help with this initial distrust. The rise of tablet computers and very cheap laptops that kids can curl up with in reading nooks may ease some of the visceral reaction against seeing kids lined up in regimented rows of monitors. Lots of laptops sprawled across classroom tables looks like your average Starbucks, rather than a cube farm.
Moreover: the Silicon Valley titans who send their children to a screen-less Waldorf-style school are able to train their children in modern technology in lots of other ways. Lower-income parents may be much more anxious for their children to learn in school the skills necessary for success in the modern economy. “I was a little bit surprised at how much blended learning appealed to poor minority parents in San Francisco,” says Scott Hamilton of Seton Education Partners. “They’re not well versed in digital learning and curricular offerings, but they know that computers are central to their children’s employability and life as adults.”
As for the changing role of teachers, “The feeling I have as a parent is I’m absolutely not willing to relinquish that teacher connection. Teacher as a mentor, teacher as a guide—that’s totally crucial, almost more than ever,” says Heather Staker, whose children attend Acton Academy in Austin. But with online instruction in core subjects freeing up teacher time, her children get more teacher attention, and get to learn how they learn best. “It’s been exciting to see my own kids find their passions,” she says. One daughter discovered she’s got a knack for math and is now soaring ahead. “I don’t think she would be unleashed that way in a traditional environment,” Staker says. That’s worth letting go of nostalgic fantasies of chalkboards and dusty textbooks.
Bottleneck 3: Misguided Policy
Education policy is often the object of a tug-of-war among different interest groups. Scads of local, state, and federal policies create obstacles for blended learning, even when that’s not the explicit intention. Many of blended learning’s first adopters have been charter schools rather than traditional district-run schools, simply because blended instruction requires fresh and flexible thinking among administrators and teachers, and that is more common at charters. But even within the charter-school sector there are serious legal and administrative obstacles:
- Some states have limits on the overall number of charter schools that block further growth.
- Some states regulate charter schools in ways that inhibit innovation, such as blocking for-profit operators or cyber charters, and demanding particular teacher-to-student ratios, even in online schools.
- Many charter schools receive far less than the amount of funding given to their district counterparts.
- Some states don’t adequately assess charter schools, allowing low-quality operations to undermine support for more rigorous institutions. (“We let a thousand flowers bloom,” says Gisèle Huff of the Hume Foundation, “and half of them were weeds.”)
Even policies that sound quite reasonable can thwart deeper thinking about education, and may limit the growth of blended learning over time. Take rules on class sizes. According to Bryan Hassel of Public Impact, 36 states have some sort of limit on class size. These rules seem well-intentioned, but they create two problems. First, they apply equally to all teachers, regardless of quality, even though 28 students assigned to a teacher in the top quartile will likely do better than 22 students assigned to a teacher in the bottom quartile. Second, if you can’t increase class sizes, then blended-learning programs are doomed to remain small and uneconomic. If you have all the start-up costs associated with the technology but can’t capture the productivity and quality gains associated with increased class sizes led by a smaller universe of superior teachers, then you’ll just have more expensive schools, not better ones.
The list of policies made for a different era goes on and on. Teacher licensing requirements can inhibit blended schools. Some union contracts present intractable obstacles to advanced digital learning. State textbook procurement processes often make no sense when content can be delivered digitally and updated constantly. Annual tests whose results are delivered far too late to actually affect instruction can just waste time in a universe where students are assessed constantly.
Susan Patrick of iNACOL says, “A huge focus of our time is on policy—policymakers asking us what are the barriers in my state. It’s nice that we’re a membership association so we can reach out to both schools and school districts and ask ‘What are you facing as barriers? What’s really happening?’”
Perhaps the most reasonable-sounding but misguided policy is that in many states, school funding is based on kids spending a certain number of hours and days at school (e.g., six hours a day for 180 days). This makes sense for ensuring a standard amount of instruction time. But this “seat time” concept is a blunt instrument for measuring learning. A school that has children passively watching TV each week is compensated the same as a school that manages to move children through and above grade levels in a lesser amount of time.
“I’m convinced that key elements of online and blended learning cannot thrive under the constraints of today’s systems of education governance, finance, and choice,” says Chester E. Finn Jr., president of the Thomas B. Fordham Foundation. “This technology has the potential to revolutionize the entire education delivery system—but the system has plenty of vested interests that will do all in their power to keep that from happening.”
Education isn’t the only field that has struggled with counterproductive public policy. In medicine, payers have for many years compensated providers for procedures carried out, rather than paying for good results. A hospital with a high rate of re-admittance due to complications has been able to bill more than a hospital that gets people home quickly and keeps them there. This creates few incentives for quality care.
Imagining different ways of compensating schools is fascinating. Would it be possible to reimburse partly based on mastery? On how much students improve compared to their starting point? What would school look like if the knowledge necessary to be a high-school graduate—or even to complete a particular grade—was broken down into specific units, and then students needed to demonstrate that they had accomplished each one? Any organization that moved the student to mastery, whether in the 1,080 hours students usually sit in chairs each year, or in more time or less, could be paid for accomplishing that.
A broader funding question is how online course providers should be compensated. In current blended-learning models, schools generally select software to be offered school-wide, and purchase it out of school funds. What about when students take pure online courses that feature teachers located elsewhere? Some policies make even this partial adoption of technology difficult. In California, notes Patrick, students can only enroll in online courses offered by neighboring districts—as if the internet doesn’t work over long distances.
Some districts are willing to pay an online provider for a course the school doesn’t offer itself. That raises another interesting question. What if a school does offer a course, yet students or parents want to enroll in an online alternative?
Well-developed online learning introduces possibilities of “school choice” on a course-by-course basis. The most enlightened idea would be for funding to follow the child wherever she goes, right down to the individual course level. A portion of the child’s education allotment from the state could go to the online operator, and a portion might also go to the bricks-and-mortar school where the child is sitting, to cover the cost of their utilities, adult supervision, and so forth. Good policies are clearly possible, if political and regulatory obstacles can be surmounted.
The process of figuring all this out is going to slow the adoption of digital and blended learning. Some states, though, are trying innovative approaches. In 2005, New Hampshire became the first state to eliminate the Carnegie unit (the standard time-based measurement for learning). New Hampshire now requires schools to grant high-school credit based on competency rather than seat time.
Philanthropists can play a big role in changing misguided policies. Foundations including the Hume Foundation, the Gates Foundation, and others have supported iNACOL. Others have supported the Foundation for Excellence in Education (run by former Florida Gov. Jeb Bush) and the Alliance for Excellent Education (run by former West Virginia Gov. Bob Wise). Together, the two former governors chair Digital Learning Now, an organization that advocates for sensible digital-learning policies.
Frequently, says Wise, what’s needed most is “clearing out the underbrush” of policies that no longer make sense. “Often, legislatures don’t do technology well. In the time it takes to reach a compromise, the technology has already advanced two to three generations.” Wise warns that “the decisions made in the next two years will set up education for the next two decades.” To encourage good policies, Digital Learning Now rates states based on how they compare to their peers. Over time, sunlight and public exposure of this sort could alter policies that discourage 21st-century learning.
Bottleneck 4: Inertia
Perhaps the biggest barrier to the growth of blended learning is that public education by nature tends to be glacial in its rate of movement and change. The hundreds of billions of dollars spent nationally on public schools have created entrenched lobbies dedicated to protecting the status quo. More neutrally, schools are the heart of many communities, and they have long traditions that people naturally protect. Sometimes educational innovation is great, but often it is just faddish. Says Michael Horn of the Innosight Institute, “There are a lot of people jumping in and doing blended learning because it’s the cool thing to do right now, without giving a lot of thought to why or what problem they’re solving.” A bad precedent could sour the movement quickly.
So, plenty of schools cling to a “wait-and-see” kind of attitude. That can be frustrating—particularly given the complacency that causes many communities to imagine their local schools are just fine, even while they recognize that the national averages are profoundly mediocre. “Parents sometimes see the flowers blooming around the flagpole out front and the new team uniforms and assume, ‘Man, our school must be kicking it!’ But they’re missing what’s important,” warns former Arizona schools head Lisa Keegan.
If the point of schools is to produce great educational outcomes for students, then many of today’s public institutions have deep, systematic inadequacies. On the other hand, if you view big school systems as job programs for adults in the community, as places to keep kids quiet and safe during custodial periods while parents work, then most are doing fine, and any change to established means of operation—particularly one involving fewer jobs and more accountability, which might eventually be the case with blended learning—is likely to be treated as an existential threat.
Joe Williams, the head of Democrats for Education Reform, says that within teacher unions “the slippery slope argument comes up. If you allow students to start taking classes online without being in a school, at some point you’ll have fewer teachers—so this is the end of the world.” Reform, he notes, “is a lot more difficult than I ever imagined. There’s still a clock-punching mentality.”
Sometimes it takes a literal act of God—as happened when much of New Orleans was wiped off the map during Hurricane Katrina in 2005—to up-end the status quo. With many schools under water, reformers came in and started opening new programs within days. Now, more than 80 percent of New Orleans children attend charter schools, some of which—like FirstLine schools—are experimenting with blended learning. Test scores are slowly rising. With this level of charter market share, actual market discipline is rearing its head, and the state board of education is forcing ineffective schools to close.
Though change is hard, philanthropists who work with younger teachers believe that generational change will aid progress in this area. Williams notes that “there are groups of younger teachers who are willing to keep their minds open for these kinds of changes.” Technology itself is neutral, and plenty of teachers now earn their own master’s degrees through what are, in essence, blended-learning programs.
Young teachers who want their elementary-school charges to hear a song will naturally pull up YouTube to find it. Internet access “is water to them,” says Caprice Young of the Laura and John Arnold Foundation. So the battles of today may not be the battles of tomorrow. The American Federation of Teachers, for instance, is working on creating a national program that would help teachers share lesson plans based on the emerging Common Core standards, recognizing how technology can make teachers much more effective. “The union leadership wants to find ways to be relevant with younger teachers,” says Williams.
Philanthropists who would like to be helpful in this process need to find ways to work with anyone interested in using technology to improve student outcomes, even if they have other disagreements. Building broad, democratic support for digital forms of learning is essential. Says Alex Hernandez of the Charter School Growth Fund: “Where we’ve seen innovation really take off is where there’s been a well-developed ecosystem of folks who want to innovate and people who want to support them. . . . Unless there’s a community of people excited about a reform, change just turns into a mandate, and mandates never work out.”
How do the Boy Scouts know a young man has mastered the science of first aid? The organization grants him a merit badge when he’s demonstrated competence in certain skills. He can display the badge proudly, and all other scouts know what it means: this boy can dress your wounds in a jiffy.
The idea of “micro-credentialing”—a fancy word for merit badges—is gaining steam in education. The bachelor’s degree, for instance, shows that you’ve attended and passed a certain number of classes, but it doesn’t necessarily indicate mastery of any particular skill per se, in a way that’s standardized across universities. Employers staring at a pile of résumés find it difficult to figure out just on the basis of a bachelor’s degree who will make a good employee. The tendency is consequently to hire people who’ve completed internships (often unpaid), who have gone to brand-name schools, or whose relatives and friends already work for that employer. None of this advances the cause of equity or a competitive labor market.
Merit badges could fill that gap. The John D. and Catherine T. MacArthur Foundation sponsored a contest in 2012, along with software-maker Mozilla, to come up with digital badges for particular skills that organizations want to assess. Sal Khan reports that the Khan Academy is working on a merit badge for college algebra and one for remedial (or “developmental”) math. “Students don’t like having to pay tuition to take developmental math,” he says, so brushing up with Khan Academy and then being able to demonstrate mastery could be a way around an otherwise costly requirement.
Mastery-based credentialing fits well in the world of information technology, says Khan. The computer-programming industry already is almost ignoring credentials like college degrees and asking candidates instead “What can you do?” Khan says he is surprised that no one has come up with a truly credible micro-credentialing system before now. “The old way of assessing is just broken,” he says. Any time he talks about merit badges, “almost everyone gives these huge nods of approval.”