In a darkened classroom at Francis W. Parker Charter Essential School in Devens, Mass., 15-year-old Tom Grigglestone is giving a PowerPoint presentation of what he’s learned in math this past semester—and how he’s applied this knowledge to a project he designed.
“I can predict where the NASDAQ will be when I know where the ‘Footsie’ has ended up,” he says, referring to the FTSE, an index of the 100 biggest companies on the London Stock Exchange.
He takes his audience through a series of slides that explain how to find the correlation between two random sets of data by using simple linear regression—pretty advanced stuff for a high school sophomore. In this case, his “audience” is just one person: his teacher, Nathan Soule, who scribbles notes on a sheet of paper as Tom is talking.
Tom is practicing for an exhibition, which the school calls a “gateway exercise,” that he must complete before advancing to the next grade level—like a graduate student’s oral examinations. Parker’s gateway exercises are a classic example of performance-based assessment, in which students show their understanding not by filling in bubbles on a standardized test but by producing actual work—an essay, a lab report, a presentation, a portfolio, or some other demonstration of competency.
Performance assessment is what teachers do every day when they grade students’ projects and assignments, but often this work is not part of the high-stakes system that determines whether students are ready to graduate—or whether schools as a whole are making progress.
For a while in the 1990s, that was starting to change, as states like Connecticut, Nebraska, and Wyoming were developing large-scale performance assessment systems. But the dawning of No Child Left Behind “pushed aside” these efforts, because it was too costly for states to include performance assessment in their statewide accountability systems under the law, said Joan Herman, director of UCLA’s National Center for Research on Evaluation, Standards, and Student Testing.
Now, the tide is turning again.
Buoyed by the move toward common standards, as well as a federal directive to use “multiple measures” of student success, two consortia of states are designing next-generation assessments that could drive more meaningful learning in schools nationwide, beginning with the 2014-15 school year. Rather than take a strictly multiple-choice test or write answers in a blue book, students will sit in front of computer screens several times a year answering questions online.
Supported by federal Race to the Top money, this computer-based testing will allow state officials to design more rigorous high-stakes exams that also include performance-based tasks, measuring students’ abilities on a large scale in ways that a simple pencil-and-paper test cannot.
The Race to the Top exams are “putting performance assessment back into assessments that count,” says Herman, who calls this development “encouraging.”
The change is welcome news, agrees Stuart Kahl, chief executive officer of Measured Progress, which develops customized assessments for schools, districts, and states.
“International tests suggest that we’re not declining; other countries are just moving ahead of us,” says Kahl, whose company is guiding the development of performance-based tasks within the tests being created by the SMARTER Balanced Assessment Consortium, a group of 28 states.
As other nations put more emphasis on performance-based assessment, Kahl says, “there’s a realization that we should be doing this, too.”
Benefits of performance assessment
For advocates of performance assessment, it’s a no-brainer why states and schools would want to use this technique more frequently: Imagine if you didn’t have to take a driver’s test before getting your license, leaving the assessment of your driving skills to a multiple-choice exam instead.
Just as a driving test is a practical assessment of whether you can handle a car on the road, performance-based assessments are superior tools, their supporters say, for showing how well students have learned the higher-order thinking skills necessary in the Information Age—such as the ability to analyze, synthesize, and evaluate information.
That’s obvious from watching Tom Grigglestone rehearse his gateway presentation at Parker Charter School.
With Soule looking on, Tom describes how his gateway project stemmed from the question: “How can I use math to make money?” He notes that linear regression can help analysts predict the future—which is an important skill for stockbrokers.
Tom says he began his project by wondering whether there was a correlation between the Dow Jones and interest rates. He pulled down 84 separate data points from Yahoo! Finance and created a scatter plot to find out. Using linear regression, he discovered there was only a 40-percent correlation between these inputs—not nearly strong enough to be reliable. So he experimented with various data combinations, until he discovered a 77-percent correlation between the NASDAQ and the FTSE.
That’s useful, he explains, because the London market closes five hours before the U.S. market—and if you knew how the FTSE performed that day, in theory you’d have about a three-in-four chance of being correct if you bet on a similar performance from the NASDAQ.
Soule is Tom’s assessor for the gateway process. The event itself is more of a celebration than an actual assessment; all the heavy lifting has been done before, in class, as Tom has been working on his project and discussing it with Soule. So, this practice run is intended more to give Tom feedback on his presentation than to evaluate his work.
“You’re going to have a slide on this part, right?” Soule asks at one point. Later, he tells Tom the presentation was really good, but a little short: 20 minutes, when it should be 30 to 45 minutes long.
Anyone watching Tom’s presentation would clearly see that he understands not only how to find correlation using linear regression, but also why this knowledge is significant and how it applies outside of school.
Because performance assessment engages students in an activity that ultimately leads to a task or product that can be scored, students tend to go “way beyond the things they learn in class,” Soule says. The result is a better understanding of students’ skills by their teacher, but also a keener knowledge of the topic by the students themselves.
Performance assessment gives students “the chance to go deeper into their learning,” says Sue Massucco, arts and humanities domain leader at Parker Charter School. “They get to know their content deeply, but they also get to activate their minds—which will last a lifetime.”
A framework for success
The inclusion of more performance-based tasks in the new Race to the Top assessments that states are creating is significant, Herman says, because it means schools are now more likely to embed these kinds of activities into their own curricula.
Studies from Herman’s organization suggest that “teachers really heed the signals sent” by high-stakes exams, she says. In other words, they often “teach to what’s on the test—and the curriculum tends to become test preparation.”
Christina Brown, director of the Boston-based Center for Collaborative Education’s Building Quality Performance Assessments (QPA) initiative, sees the reemergence of performance assessment as a critical opportunity for schools.
“There’s so much to be gained by teachers scoring student work,” she contends, noting that educators are in the best position to gauge their students’ abilities. “If teachers can’t [be trusted to] score work reliably, how can they teach the skills we require of students?”
But to take full advantage of this opportunity, Brown says, school leaders “have to do a better job of teaching teachers how to implement performance assessment with technical quality”—that is, with fidelity and consistency.
Brown’s QPA initiative is working with 20 New England schools and districts, including Parker Charter School, to help them do just that. The group also is releasing a report this month, called “Quality Performance Assessment: Harnessing the Power of Teacher and Student Learning,” that summarizes the lessons it has learned.
Co-authored by Brown, the report defines high-quality performance assessments as “multi-step assignments with clear criteria, expectations, and processes that measure how well a student transfers knowledge and applies complex skills to create or refine an original product.”
The report, which will be available at www.qualityperformanceassessment.org [1], also offers a framework for delivering high-quality performance assessments. Here are the framework’s three elements:
• Aligned instruction. All students need instruction based on college and career readiness standards that are accessible to students’ learning strengths and needs.
• Task design. Designing valid assessments involves determining appropriate levels of content and cognitive complexity, setting clear criteria for success through a detailed scoring rubric, and ensuring accessibility for all students.
• Data analysis. By examining student work, educators can infer important information for planning future instruction and assessment.
At the center of this framework is “authentic student learning,” which is defined as “learning that is meaningful to students and measures complex skills … that are transferable to new situations.” Learning is most authentic, the report says, when it gives students opportunities for ownership of their studies and decision making in real-world situations.
The need for a common language
For performance assessments to be valid and reliable, the tasks’ scorers must agree on what a particular rating means, Brown says—and they must score each assessment the same way.
Think of teachers as umpires, she explains: They need to understand and agree on “where the strike zone is.”
Reaching this agreement can be hard work. It requires teachers to work in shared communities of practice, discussing what high-quality work looks like and developing a common language to define it.
At Parker, this is accomplished by having common planning time that is devoted to assessment.
Each assessment is co-designed by a team of six teachers who work in that domain, or area of knowledge. The assessments are backwards-designed, meaning teachers start with the goals they want students to achieve (based on the Common Core standards and the Massachusetts frameworks) and then design a series of tasks around these goals.
It can be a challenge for six people to agree on anything, Massucco acknowledges.
“Collaboration is hard,” she says, noting that in a traditional school setting, teachers can simply choose to close their doors to colleagues—but “not so here.”
Still, she wouldn’t trade her experience for anything. “A bubble test doesn’t allow [for] revision,” she says, and yet that’s a key feature of performance assessment. Students receive valuable feedback on their work, just like they’d get from their employer in the real world—and they can revise, retry, and perfect it until they master the task.
Getting stakeholder buy-in
To ensure the success of its performance-based assessment model, Parker has a unique schedule that gives teachers two hours of common planning time per day.
“When other teachers hear that, they say, ‘Oh my gosh, it’s Shangri-La,’” Massucco says.
But the Pentucket Regional School District, north of Boston, proves you don’t have to operate a charter school to pull off performance assessment well—and you don’t have to blow up your schedule, either.
Two years ago, Pentucket began a process to redefine the skills students should be able to demonstrate before graduating ready for college or a career.
District officials wanted to assess these skills with the use of student portfolios in grades four, six, eight, and 11. To make sure parents and teachers were invested in the project, the district held multiple small-group sessions to define these skills and the learning that would accompany them—and all stakeholders were invited.
“The fact that it’s a community-created document puts a lot of weight behind it,” says Assistant Superintendent Bill Hart.
From these meetings evolved Pentucket’s five Habits of Learning: thinking, communication, collaboration, independence, and creative exploration. These five Habits are infused throughout the district’s core curriculum.
“It drives me crazy when people talk about these skills as soft skills,” Hart says. “We have to think of these 21st-century skills as content that should be taught—and then give students feedback on their progress toward meeting these skills.”
As was the case at Parker, educators needed time to develop a shared vision for how to teach and evaluate these skills. In the project’s first year, Pentucket gave its teachers 20 early release days—90 minutes of extra time every other week—for this planning and collaboration to occur. This school year, teachers will have 16 early release days.
Teachers also could attend voluntary summer sessions, for which they received either in-service credit or a stipend. Although these sessions were not mandatory, the district still had at least 50 percent of its teachers participate, Hart says.
And though there was some pushback from staff, he says, spelling out clear expectations helped—as did forcing the issue.
“To bring about change,” Hart says, “you have to be courageous enough to say: This is what we think is best for students, and this is what we’re going to do.”
A ‘performance-driven world’
Teachers and administrators at both Parker and Pentucket agree: When parents see an exhibition of their children’s work, it’s a powerful moment.
“I can’t tell you how many parents left in tears,” Hart says, because they were proud of how much work their children had done—and moved by how well their children were able to articulate it.
Performance assessment resonates with parents, he adds, because they “want their kids to be able to navigate what is a performance-driven world. … I’ve never heard a parent say it’s important to memorize the quadratic formula.”
Massucco shares a similar thought: “Performance assessment is a great thing. It’s what people do in their lives. I’ve never taken a test in my job.”
Before NCLB introduced an era of greater accountability in schools, there wasn’t a lot of control over the quality of performance assessment, Kahl acknowledges, and that led many critics to deride what they believed was its “loosey-goosey” approach.
But aligning performance-based assessment to common standards and ensuring its reliability, he says, will go a long way toward easing critics’ concerns—and he sees a place for performance assessment alongside other measures of student achievement.
“Is it worth the additional time and expense?” he asks. “I think it is. But you have to account for this additional time and expense.”
Brown remembers the earlier criticism well, and that’s one reason she’s so passionate about her center’s work in assuring high-quality performance assessment.
“I feel like this is the window where we either get this right—or we screw it up for a really long time,” she cautions. “And that would be sad. We’d lose a whole new generation of students.