Around the country, science instruction is changing—students are being asked to make models, analyze data, construct arguments, and design solutions in ways that far exceed schools’ previous goals.
That means science testing, of course, needs to change as well.
Students “have got to show us how they know, not just what they know,” said James Pellegrino, a co-director of the Learning Sciences Research Institute at the University of Illinois at Chicago and an expert on assessment.
Yet considering federal requirements around science testing, and states’ logistical, technical, and financial limitations, putting a new, performance-heavy state science test in place is no easy task.
Of the 18 states now using the Next Generation Science Standards, which were released in April 2013, only Illinois, Kansas, and Nevada, as well as the District of Columbia, have moved completely from their previous science tests to ones that align to the newer “three-dimensional” benchmarks. Illinois and the District of Columbia were the first to take the leap, putting an operational test in place in spring 2016. Illinois did so especially quickly to comply with federal reporting requirements, designing a new test in just six months—a move some experts have questioned.
“I was really kind of surprised they went forward with the assessment as soon as they did,” said Peter McLaren, a former state science supervisor for Rhode Island, who now consults with states and districts on NGSS implementation.
Most other states that have adopted the NGSS are taking things a bit more slowly, aiming to start operational tests aligned to the standards in 2019 or later, with a few, including Kentucky, an early adopter, aiming for 2018. About a dozen more states are using standards based on the same framework as the NGSS, and many of those are on similar schedules for implementing large-scale tests.
Federal Pressure
Complicating the changeover to new assessments is the fact that federal law (both the No Child Left Behind Act and the new Every Student Succeeds Act) requires states to test students in science at least once in grades 3-6, 6-9 and 10-12 each year. Consequently, many states are giving tests aligned to their previous state standards, despite the fact that students may be learning the updated standards. And at the same time, they’re pilot- and field-testing the new assessments.
That’s the case in Washington state, which adopted the Next Generation standards in 2013. Students are currently taking exams based on to the 2009 standards. NGSS-aligned exams are being designed and field-tested now and will go into place next year.
“We were a little earlier adopting than those states around us [including Oregon and California], so that puts a little more pressure on us to get going,” said Cinda Parton, the director of assessment for the Washington education department. With the new tests, “I think we’ll be on a cycle of continuous improvement.”
Wisconsin is piloting NGSS-type items on its current science assessment. The state is not an NGSS adopter, but it’s locally controlled, and about 80 percent of districts there are using the standards, said Kevin Anderson, the science education consultant at the Wisconsin education department.
“We’re trying to do our best to live in both worlds, realizing that moving forward we want a test that’s more in line with current research in science education.”
California has taken its own tack, asking the U.S. Department of Education for permission to skip its yearly testing and instead only give pilot NGSS tests for 2017 and 2018. That means students wouldn’t get scores for those two years.
“There was no desire on any front to continue with the old standards and the old test,” said Keric Ashley, the deputy superintendent for the California education department. And “to slap together a test to meet the requirements didn’t seem to be the right approach.”
The department refused the state’s request last December. But California, somewhat boldly, went ahead anyway with its pilot tests this spring. And now, officials say, the federal government is reconsidering.
“We are in conversations about a one-year waiver,” said Ashley. “Based on those conversations, we’re hopeful it will be approved.”
Promise of NGSS Tests
The Next Generation Science Standards outline what K-12 students need to know about physical, life, and earth and space sciences, yet in a more complicated format than previous state standards. The standards were developed by 26 state partners in collaboration with science, science education, higher education, and business and industry groups.
Each standard has a performance expectation, which lays out what students should be able to do. For instance, they may need to make a model, conduct an investigation, or interpret data. Standards also have three other “dimensions": disciplinary core ideas, science and engineering practices, and crosscutting concepts. These describe the facts students need to know, what they should do to act like a scientist or engineer, and how the facts connect to other fields of science.
Previous standards tended to present the facts and aspects of scientific reasoning, such as what makes for a good experiment, explained Pellegrino, who is working on a National Science Foundation-funded project to create performance tasks for classrooms using the NGSS. And the associated tests tended to assess students on the facts and inquiry separately.
The point of the NGSS, alternatively, is to fully integrate knowing and doing in the classroom.
“This now is a much deeper challenge in terms of what we’re expecting kids to show us,” Pellegrino said.
Large-scale tests aligned to the new standards should have sets of interrelated questions—known as clusters—that refer to a single “stimulus,” such as a graph, an image, or an animation of an experiment, according to a 2015 report by the Science Assessment Item Collaborative, convened by the Council of Chief State School Officers in partnership with WestEd, which develops assessments.
The tests should ideally be given online to allow for interactivity—so a student can, for instance, simulate collecting data or running an experiment. And they should be able to measure students’ skills on a continuum of learning.
The tests being created so far do have more bells and whistles than previous ones, but students won’t mistake them for video games or authentic lab work.
In many cases, the test-delivery platforms will look similar to those being used for math and reading tests aligned with the Common Core State Standards.
In a sample middle school item from the American Institutes for Research, one of several vendors working with states to produce NGSS-aligned tests, students see an animation of Jupiter’s moons orbiting the planet.
They’re given a tool to measure distance within the animation and shown data on each of the moons by name. Then they’re asked a series of questions that allow them to determine which moon is which.
“You’re asking kids to not just exhibit knowledge but to apply and use that knowledge,” said Jon Cohen, the president of AIR assessment. “We’re able to do all kinds of cool stuff.”
Illinois Out in Front, Amid Scrutiny
Given that NGSS test questions are labor-intensive to create, some states plan to share items with one another.
Illinois, which did not administer science tests in 2015 and received a warning from the federal government for being out of compliance, needed to craft a test quickly last year. The state made a deal with the District of Columbia, which already had a bank of items, to be able to use those test questions and return the favor down the road.
The state produced approximately hour-long tests for 5th and 8th grades and for biology classes. The tests each had about 20 items and included multiple choice and constructed response, or short-answer, questions.
The state also had to set up a variety of platforms for managing and delivering the test, Daniel L. Brown, the division supervisor for assessment for the Illinois state board of education, explained.
“Ultimately, in my position, I’d love to have two years to be able to build this system. It’s not ideal to have to do it in six months because you introduce all sorts of risk,” he said. “We tried to satisfy a requirement and at the same time build something for the future.”
Some experts see Illinois as a cautionary tale. “They are trying to make use of assessments that were hastily assembled to try to align with the Next Generation Science Standards for which the evidence is lacking about how valid they are,” said Pellegrino. “They had no money and no time. … The Illinois case illustrates the exact problem all states are going to have if they try to do this too quickly.”
Brown said the first year of testing had some technology problems, but he thinks the test lined up with the three-dimensional nature of the standards. And this year, test administration has gone much more smoothly, he said. However, because of delays in the state budget, results for even the first year’s test have yet to be seen.
“As of today, we’ve got about 1.4 million items that have to be hand-scored,” Brown said last month. “We’re about 10 percent in.”
The District of Columbia is going through benchmark-setting for its scores and plans to release results from its first test in June.
In Nevada, which began NGSS-aligned tests this spring, officials said they know other states will be watching their outcomes.
“It took courage to move in this direction,” said Peter Zutz, the administrator of assessment, data, and accountability management for the state’s education department. “All eyes are on Nevada.”
Officials in California, for their part, expressed relief at having some leeway in starting their official tests.
“Had this been a different subject matter, and different standards, it might have been possible to speed up the development of the test,” said Ashley, the deputy superintendent there. “This assessment could not be built in a short period of time.”
