69传媒

One classroom, 109 children.

For the next 40 minutes, half the 4th graders here at Rocketship S铆 Se Puede Academy, a charter elementary school in a low-slung building nestled below Highway 680, will be split between the language arts and mathematics sections of the large room, working with two-dozen similarly skilled classmates and a credentialed teacher. A few will sit at a small table with the class鈥� lead teacher, receiving in-depth reading instruction targeted to their individual strengths and weaknesses. And roughly half the children will settle in front of a laptop, practicing math problems with help from an animated penguin named JiJi.

Welcome to the new 鈥渇lexible classroom鈥� from blended learning pioneer Rocketship Education鈥攐ne of the country鈥檚 fastest-growing charter school operators, with plans to open as many as 51 more schools in nine regions over the next five years.

The Tension Between Scale and Quality

As the Rocketship charter network has added schools and students, the percentage of its students scoring proficient on California state tests has steadily declined. The organization aims to grow aggressively, with plans to add as many as 51 schools over the next five years.

SOURCES: Rocketship Education; At the request of Education Week the California Department of Education calculated the total percentage of all tested students in Rocketship schools who scored proficient or above on California Standards Tests in each of the past five school years. Only those students who met al CDE criteria for inclusion in state accountability reports were included.

Many look at Rocketship鈥檚 new approach, which represents a significant departure from the blended learning model that helped launch the Bay Area nonprofit to national prominence, and see the future.

But the story behind the organization鈥檚 flexible classrooms is also a cautionary tale about the belief that technological innovation can fuel rapid school expansion without compromising quality. Although test scores have steadily declined as the network has added schools and students, Rocketship has maintained its voracious appetite for growth. Rather than resolve that tension, the new flexible classrooms have, by Rocketship鈥檚 own admission, further strained the organization and exposed underlying problems glossed over during the group鈥檚 ascent.

Some Rocketship leaders, for example, now acknowledge that their original blended learning model鈥攚hich powered the organization鈥檚 initial growth, to nine schools and 5,200 students, before its impact could be rigorously studied鈥攎ay be more effective at teaching students to follow directions than to think for themselves.

Red flags also have come up around finances: Documents and interviews make clear that the new, flexible classrooms were originally devised as part of an audacious plan to cut staff, save $200,000 annually per school, and redirect the funds to help start up new Rocketship schools鈥攁 strategy that proved too aggressive, even for Rocketship, and has since been dialed back.

And the on-the-ground reality has been messy, too. A sweeping experiment with flexible classrooms during the 2012-13 school year resulted in sharp networkwide test-score drops and dissension among the organization鈥檚 rank and file.

Rocketship CEO and co-founder Preston Smith staunchly defends the organization against critics who accuse it of pushing too hard to expand blended learning鈥檚 boundaries.

But even Mr. Smith concedes the network is at an 鈥渋nflection point.鈥�

鈥淲e need to figure out the flex model and get much more grounded on how we can consistently realize the levels of achievement we expect,鈥� he said. 鈥淚t puts us in this conundrum of do you continue to focus on innovation, and try to get it right, or do you actually start to push toward scale?鈥�

Rethinking 鈥楽tation Rotation鈥�

Founded in 2006, Rocketship鈥檚 motto is 鈥渞ethinking elementary school from the ground up.鈥�

In pursuit of that lofty goal, the organization has relied on technology to accelerate student achievement, while cutting labor costs. For years, schools in the network have used the 鈥渟tation rotation鈥� model of blended learning, with students cycling each day between about six hours of traditional classroom time and two hours of computer-assisted instruction in 鈥渓earning labs.鈥� That model, which helped give birth to the blended learning movement, has allowed Rocketship to replace one credentialed teacher per grade with software and an hourly-wage aide, freeing up $500,000 yearly per school that can be redirected to other uses.

The approach generated impressive test-score results, particularly in math: Rocketship鈥檚 early academic results far outpaced state targets, ranking its schools among the best in California when it came to serving English-language learners and students from low-income families, who represent the overwhelming majority of Rocketship students.

Mr. Smith said those results led to immense parental demand, prompting the organization to try to grow as quickly as possible.

This fall, Rocketship opened a new school in Santa Clara County, in California鈥檚 Silicon Valley, as well as its first school in Milwaukee. A school in Nashville, Tenn., and another in the Bay Area are on tap to open in the fall. By 2017-18, Rocketship aims to serve more than 25,000 students in 60 schools across the country, according to board documents.

But as the network has grown, scores on California state tests have trended downward. According to an analysis performed at the request of Education Week by the California Department of Education, the number of Rocketship students scoring 鈥減roficient鈥� or above in English/language arts has plunged 30 percentage points over the past five years, to 51 percent, while math proficiency rates have dropped more than 14 points, to 77 percent.

Lynn Liao, Rocketship鈥檚 chief programs officer, said the organization has also received troubling feedback on how students educated under the original blended learning model fare in middle school.

鈥淎necdotal reports were coming in that our students were strongly proficient, knew the basics, and they were good rule-followers,鈥� Ms. Liao said. 鈥淏ut getting more independence and discretion over time, they struggled with that a lot more.鈥�

Like other blended learning practitioners and advocates, Rocketship has now come to believe the station-rotation model is too rigid and leaves too wide a gulf between the computer-based learning labs and traditional teacher-led classrooms.

鈥淭here鈥檚 [now] a general consensus that...station-rotation can produce impressive student outcomes within the narrow band that state testing represents,鈥� said Andrew J. Calkins, deputy director of Next Generation Learning Challenges, a Washington group that backs blended learning. 鈥淏ut by itself, it鈥檚 insufficient to bring students to the level of skills and knowledge required for 21st-century success.鈥�

Complicating matters for Rocketship were weak school budgets and the possibility of further cuts in California state funding鈥攂oth of which threatened the organizational growth slated to begin in fall 2013.

Rocketship鈥檚 response, hatched in mid-2012, amounted to doubling down on the formula that first made it a national phenomenon: To bolster academic quality, generate more efficiencies and continue growing, the group sought to leverage the power of technology even further.

Three months into the 2012-13 academic year, Rocketship began knocking down classroom walls in order to quickly pilot the new, flexible classroom model.

A Rocky Rollout

A timer inside S铆 Se Puede Academy鈥檚 new, wide-open 4th grade classroom turns to zero, and the three teachers in the room give each other quick nods. One turns on bouncy pop music, the soundtrack that accompanies 鈥渞otations鈥� in Rocketship鈥檚 flexible classrooms. Like that, all 109 children in the room are on the move.

Within three minutes, four separate groups of students are settled in new seats, engaged in four very different lessons.

The seamless rotations now are a far cry from the chaos of the 2012-13 school year, said Principal Andrew Elliott-Chandler.

During that year, when Rocketship was piloting its new, flexible blended learning model, the changes at S铆 Se Puede began in 2nd grade in November. By December, the entire school was testing an alternative version of the model that had flexible groups of students rotating among separate classrooms. In April, Rocketship changed things up again, moving the 2nd, 4th, and 5th grades to the flexible spaces that remain in place today.

鈥淜ids like to know the rules of the game,鈥� Mr. Elliott-Chandler said. 鈥淲hen you try and change midyear, it鈥檚 very challenging.鈥�

On paper, rapid overhauls of the basic building blocks of school, from classroom space to student schedules to teachers鈥� roles, was consistent with Rocketship鈥檚 tech-startup ethos.

In practice, it meant teachers had to continually rethink their strategies for controlling student chatter, passing out homework, and nearly everything else鈥攁nd that even small classroom adjustments now required consensus among several colleagues.

Making things more difficult, Rocketship initially sought to use the new flexible classroom model to shed professional staff.

In public, the organization has historically characterized its schools鈥� student-to-teacher ratio as 27-to-1, but that figure includes tutors, who are not required to have a college degree and who make roughly $30,000 per year. In reality, the network鈥檚 ratio of students to credentialed teachers is 37-to-1.

According to board documents and interviews, Rocketship leaders initially considered using the new flexible classrooms to push that figure as high as 50-to-1. They also believed one assistant principal per school could be cut under the new model.

The result would have been $200,000 in annual savings per school, on top of the 15 percent management fee that Rocketship already charges each of its schools.

Mr. Smith, the CEO, acknowledged that the organization鈥檚 initial intent was to use that money to provide startup funds for new schools, but said that plan has since been scrapped.

Gary J. Miron, a professor of education at Western Michigan University, in Kalamazoo, who studies charter-management organizations, questioned Rocketship鈥檚 willingness to even consider such an approach.

鈥淚t is wrong to call these schools 鈥榗harter schools,鈥� 鈥� Mr. Miron said. 鈥淭he appropriate name should be 鈥榝ranchise schools鈥� or 鈥榗orporate schools.鈥� 鈥�

The rocky pilot of the new flexible classroom model also led to internal complaints.

The organization lost 29 percent of its teachers after the 2012-13 school year, badly missing its internal targets for retention.

Those who stayed remained skeptical, Mr. Smith told the Rocketship board.

鈥淭here is a consensus amongst school staff that the instructional model will not help Rocketship achieve its mission,鈥� he wrote in an August 2013 memo.

Test scores plummeted in the first year of the new model, too.

Mr. Elliott-Chandler, the S铆 Se Puede principal, summed up the problem.

鈥淭he story became about change and innovation, which is different than focusing on writing instruction,鈥� he said.

Slowing Down

For critics, there are plenty of reasons to be skeptical of Rocketship鈥檚 emphasis on quickly bringing innovation to scale, both in California and across the country.

鈥淚t might work for developing technology, but these are children,鈥� said Roxana Marachi, an associate professor of education at San Jose State University. Ms. Marachi advocates Rocketship 鈥渉alting their expansion plans and putting more resources into research on what they鈥檙e already doing.鈥�

While that appears unlikely, the organization has eased off the throttle.

Shortly after becoming CEO in March 2013, Mr. Smith limited Rocketship鈥檚 use of flexible classrooms to grades 4 and 5, where the organization found the most academic benefit. Kindergarteners remain in self-contained classrooms, and grades 1 through 3 continue with the station-rotation model.

In addition, said Mr. Smith, the flexible classrooms are no longer being looked at to generate cost savings.

And for now, the new model is also being limited to the Bay Area: Rocketship Southside Community Prep, which opened in Milwaukee in September, is not using flexible classrooms in its first year, and neither will the new school set to open in Nashville next fall.

In his August 2013 memo to the Rocketship board, Mr. Smith wrote that the rapid change last school year 鈥渢urned up pressure on schools while reducing focus and support on academic results鈥� and that 鈥渦nderlying gaps in our academic systems and teacher preparation became apparent.鈥�

In an interview, he said the organization has committed to going slower with its new model.

鈥淲e stepped back and said, 鈥極K, we think there鈥檚 something here worth pursuing,鈥� and now we鈥檙e doing it in a much more limited and focused manner,鈥� he said.

But Rocketship still faces pressure of several sorts.

In Milwaukee, for example, the organization missed its student-enrollment targets this year by a significant margin, leaving a roughly $1 million hole in its new school鈥檚 budget.

Meanwhile, blended learning enthusiasts, undeterred by such troubles, remain eager for Rocketship to remain aggressive.

鈥淭he advance guard in the field has to be willing to take something of a risk, because there鈥檚 a lack of evidence that the traditional school model has brought all of our students where they need to be,鈥� said Mr. Calkins of Next Generation Learning Challenges.

The greatest pressure, though, comes from within: Rocketship officials maintain a steadfast conviction in the power of the educational model they are developing.

Sitting in the flexible 4th grade classroom at S铆 Se Puede Academy, Mr. Elliott-Chandler watches a teacher take six students through a lesson on finding synonyms in a text. It鈥檚 high-level reading instruction the children likely wouldn鈥檛 have received as the lone advanced student in a typical classroom of 30, sitting alongside many children still learning how to sound out words.

Those struggling students will now get tailored lessons, too; Rocketship鈥檚 flexible classrooms allow for daily work with small groups of students at eight different skill levels.

Despite the painful changes of the past year, Mr. Elliott-Chandler is optimistic about the future and says he couldn鈥檛 imagine a return to the traditional public school where he started his career.

鈥淚t鈥檚 now 10 years since I started working there, and they鈥檙e still struggling,鈥� he said, the old Rocketship swagger resurfacing.

鈥淭hat鈥檚 what incremental change will get you: incremental growth.鈥�