Sound and Unsound Strategies in Science Instruction

By Sean Cavanagh — June 19, 2009 3 min read
  • Save to favorites
  • Print

Let’s say you’re a middle school science teacher tasked with improving the achievement of 8th graders through textbook lessons, lab experiments, tests, in-class demonstrations, or some combo of all of them. What lessons and techniques work best? A new study seeks to provide answers.

The study, produced by a Boston College scholar and researchers from the Educational Testing Service, identifies a number of strategies that the authors say are connected with higher test scores. They include having students read science textbooks, in which test scores were shown to increase with the frequency of reading texts. Other positive strategies included having students perform hands-on activities, write extended answers to science test questions, discuss measurements and results from lab activities, and work with other students on science activities or projects. Many of those activities are as standard as can be in classrooms, of course; others, maybe not.

Several other approaches were also found to be effective in science, when used in moderation, the study found. These included having students watch teacher-led demonstrations, take science tests, prepare written science reports, and (journalists, rejoice!) discuss science in the news. Perhaps not surprisingly, students reported that reading science textbooks, doing hands-on experiments, and participating in teacher-led activities were common in their classes.

But what strategies are associated with lower average scores? Having students give oral reports on science did not produce results, the study found. Intriguingly, neither did having them use library resources in science. Are students given too little direction with these library assignments? Are they pushed to understand scientific facts, and think like scientists?

The authors were Henry Braun, of BC, and Richard Coley, Yue Jia, and Catherine Trapani of ETS. They used statistical analyses such as hierarchical linear modeling to probe students’ scores on the National Assessment of Educational Progress, or NAEP. They drew from federal data collected from students and teachers, including information about educators’ classroom techniques, in coming up with conclusions. While the authors acknowledge that information collected from student responses can be unreliable, the overall findings “cannot blithely be interpreted casually.” They also say that “the consistency across the different analyses…suggests these finding should be taken seriously” and serve as a foundation for more research.

The authors further note that their findings are consistent with some of those made in previous reports, such as the National Research Council’s “Taking Science to School.” That study called for students to master a relatively small number of science topics, but also become adept at the processes and approaches used by actual scientists—generating and judging scientific evidence, for instance.

What kinds of teachers were most effective in raising test scores? Those with standard teaching certificates did better, but only by a slight margin, the authors found. In addition, teachers whose total experience exceeds their science-teaching experience did slightly worse. I would assume that those educators would be middle school teachers who were generalists, perhaps brought into science teaching after they’d been in the profession a while.

The report is certain to prompt debate about the balance between teacher- and student-led classroom activities, hands-on experiments and textbook learning, and so on. If you’re a science teacher, do you find these conclusions surprising? Has it reinforced or contradicted your notions of what works?

Photo by Michael Dwyer for Education Week

Related Tags:

A version of this news article first appeared in the Curriculum Matters blog.