Putting science to the test 

NCLB's drill-and-kill mentality puts new pressure on science instruction

1. All living things contain which element?

a. helium
b. sodium
c. copper
d. carbon

2. Which organ removes cell waste from the blood?

a. large intestine
b. small intestine
c. kidney
d. heart

3. Which of the following gases do plants use in photosynthesis?

a. carbon dioxide
b. oxygen
c. hydrogen
d. carbon monoxide

4. Where is most of Earth’s water located?

a. glaciers
b. oceans
c. lakes
d. rivers

5. Which of the following questions is testable in a scientific investigation?

a. Are dogs better than cats?
b. Are dogs happy when they are walked?
c. Are cats easier to take care of than dogs?
d. Are cats more active at night than during the day?

Answers: 1) d, 2) c, 3) a, 4) b, 5) d

So, how’d you do? Hopefully pretty well, since these questions were taken from an old Grade 5 California Science Standards Test. But even if you missed more than you got right, nobody’s counting.

The same can’t be said for America’s students. Beginning with the 2007-08 school year, the federal No Child Left Behind Act will require districts and states to include science in their standardized, high-stakes testing.

Science educators are generally thrilled at the prospect. They say it’s about time. But their applause is not entirely unrestrained. They have cause for concern.

“It’s a good first step, but it’s just a first step,” says Gerald Wheeler, executive director of the Arlington, Va.-based National Science Teachers Association. “There are still a lot of issues to grapple with, from what’s being taught to how science is assessed. This gets science back into the big picture, but it doesn’t solve everything.”

There’s plenty to solve, to be sure. As much as some bemoan student performance in math and literacy, student academic performance in science may be worse. In the 2000 National Assessment of Education Progress, a quadrennial federal assessment of student achievement in grades 4, 8 and 12 across several subjects, only one in four U.S. students scored “proficient” or higher in science. Valid concerns have been raised about NAEP’s reliability, to be sure, but analogous shortcomings in students’ mastery of science are easy enough to find. Less than a quarter of graduating high school seniors taking the 2003 ACT college entrance exams, for example, were prepared for college biology, and general science scores were similarly problematic.

American students’ scientific education often suffers in comparison to other countries, a reality with sobering implications: future shortages of home-grown scientists, an electorate unable to distinguish between rational and emotional argument, and a public discourse lacking a nuanced scientific perspective.

In 2002, President Bush signed NCLB into law, boldly declaring that it would change the culture and performance of education in America. Whether it has actually done so is a matter of considerable debate. The law’s provisions for accountability (specifically, what it defines as acceptable academic improvement and progress) have been the source of much confusion and conflict. Many educators and organizations, including the California School Boards Association, contend that NCLB is fundamentally flawed. Without substantive changes, they say, it’s ultimately unworkable.

The law is up for reauthorization this year. There is much talk about reworking it. But at this point, what happens next is anybody’s guess.

What seems perfectly clear right now is that, until this year, NCLB has had little positive effect upon science education. Indeed, the opposite may be true, according to Maria Alicia Lopez-Freeman, executive director of the California Science Project, a University of California-based professional development network for science teachers at all levels.

“When NCLB first came in, what happened in many [elementary] schools was that the curriculum became focused just on what was going to be tested: math and reading. Science, which had never been taught that much, became a subject without any kind of preference. It became something you taught maybe on Fridays, after reading and math. It became even more marginalized,” Lopez-Freeman says.

In that sense, the NCLB science-testing requirement is welcome. But whether it has any immediate impact is unknown, say some observers, because standardized science scores won’t be weighted. They will not affect a school or district’s overall performance evaluation.

And if the science numbers don’t count, “districts and schools don’t really have any real motivation to change the way they’ve been doing things,” says NSTA’s Wheeler. “If they’ve been poor in science they can remain poor, because poor student performances won’t really affect their scoring or standing.”

That’s not to say testing won’t affect teaching. Wheeler and others believe high-stakes testing in science will undoubtedly change the way some teachers approach the subject. They’re just not sure it will be altogether good.

Instructional models

It has become a popular principle and practice among many science educators that the best way to teach science is to allow students to learn by doing, not just by sitting and listening. It’s called inquiry-based instruction, an approach that features hands-on experimentation by students with broad, periodic instructional oversight. The idea is for students to learn by thinking through problems themselves, by making their own mistakes and discoveries along the way.

The alternative—dubbed “direct instruction”—is more traditional: a teacher presenting a highly structured, explicit lesson at the front of a classroom, with students taking notes.

Direct instruction is cheaper and faster, and it’s the way science has generally been taught for decades. Critics say it’s also ineffective. Direct instruction lessons, they assert, can easily regress into lectures that are heavy on the rote recitation of scientific facts. Students are reduced to automatons instead of cultivated as growing, developing minds.

David Klahr and colleagues at Carnegie Mellon University and the University of Pittsburgh enlivened the debate between the two instructional models in 2004, when they published a study concluding that students taught through direct instruction were more likely on average to become “experts” in designing scientific experiments—a key to developing science reasoning skills—than students taught through inquiry-based instruction. But Klahr cautioned against reading too much into the findings, saying that teachers shouldn’t teach one way at the expense of the other. The approach that works best, he said, depends upon what’s being taught.

CSTA president Gilbert agrees, observing that direct instruction versus inquiry-based instruction isn’t a case of either-or.

“If a teacher isn’t qualified to teach inquiry-based instruction, then maybe all the students are doing is tinkering with lab equipment. They’re not really learning much. But when it’s done right, inquiry-based instruction is very powerful and effective,” Gilbert asserts.

Even so, supporters of inquiry-based instruction are worried that their favored approach will not test well—and so will open itself to more criticism and possibly abandonment. But Gilbert says most existing standardized tests do not fully or accurately assess students’ science knowledge, either.

“Right now, [many tests are] mostly filling in bubbles, which is geared toward recall, but doesn’t really address higher abilities, such as critical thinking and problem-solving, which is what science is really about,” Gilbert says.

Take, for example, the questions in the pop quiz at the beginning of this article. The first four questions required just one- or two-word answers: carbon, kidney, carbon dioxide, oceans. Creating and scoring such questions is easy—the answers are either right or wrong, no ifs, ands or buts.

But probing a student’s understanding of scientific method, of how to design an experiment so that it is unbiased and stands up to scrutiny, is obviously much harder. That’s what question 5 attempts to do when it asks you to identify the testable scientific investigation. The first three answers—Are dogs better pets than cats? Are dogs happy when they are walked? Are cats easier to take care of than dogs?—are not scientifically testable. Any answer would be subjective, not based on scientific precepts like quantifiable data.

How, for example, do you empirically define “better,” “happy” and “easier?”

Only the last choice—Are cats more active at night than during the day?—resembles a real scientific inquiry; only it can be addressed and assessed through acceptable scientific methods.

Wayne Carley, the executive director of the National Association of Biology Teachers, fears that science test designers may opt for the easier path, choosing questions with simple, easily assessed answers. “My fear is a back-to-basics, backwards step,” Carley says. “It comes from what I have seen in other high-stakes testing programs. Rather than teaching essential content, teachers tend to focus on the content of the test.”

In other words, teachers will resort to direct instruction, with its emphasis on facts and memorization, because that’s what the tests will be asking for—and what students (and teachers and districts) will be judged upon.

Accountability

And being judged, says the CSTA’s Gilbert, is what it’s really all about. Becoming part of standardized testing is good. Becoming part of the test score is better. In California, the Academic Performance Index does not capture test scores in science until high school. Even then, science accounts for only 8 percent of the total high school API—a virtually negligible number.

“Right now, test weighting of science is extremely low compared to math and literacy. If you have a school underperforming in math, it’s going to face sanctions if it doesn’t make the API. The principal is going to be held accountable to the district, and the district to the state, and the state to the federal government,”
Gilbert warns.

“That doesn’t happen so much with science, and obviously we think you need to prioritize things so that science education carries some of that weight, too. This alone will help [improve science education] by affecting where schools and districts put resources, such as boosting instructional time or professional development.”

At the same time, test assessments need to be improved, says Lopez-Freeman at the California Science Project. They should be aligned with desired results. “We want kids to know facts, but we also want them to be able to think critically, which is knowledge that is harder to assess. There are ways to do that. More writing assessments, for example. There are tests in Europe that do a good job. I think we can too,” Lopez-Freeman explains.

Like all science educators, Lopez-Freeman wishes the value and importance of science education was obvious to all—and supported by all. It’s not for a variety of reasons, among them competing agendas, limited resources, a lack of vision.

The NCLB testing mandate isn’t a cure-all, say people like Wheeler, Gilbert, Carley, Lopez-Freeman and others, but at least it recognizes there’s an ailment in need of redressing.

“If testing is the only way to drive science forward, then OK,” says Lopez-Freeman. “But that’s a minimum. Science is so much more than just recounting facts, filling in bubbles. It’s a foundation for national progress and what this country is all about. Science is liberal arts for the 21st century.”

Scott LaFee is a contributing writer for California Schools.