The problem, the study concluded, is that there are too few state science assessments and the information the tests seek to measure is divorced from the goals of the science-reform movement.
“State science assessments, from a national perspective, are not a driving force in directing curriculum or instructional practices toward the standards of science-education reform,” according to the ets research team that conducted the study.
Moreover, the researchers found, “almost all content structures in the state assessments conflict with the current science-education reform goals.”
The researchers, Brian Gong, Colleen Lahart, and Rosalea Courtney Writing, said in their study, “Current State Science Assessments: Is Something Better Than Nothing?” that they found it “surprising and disturbing” that 24 states have no statewide science assessment.
While state science assessments need to be improved, the researchers argued, even flawed examinations could serve the purpose of encouraging schools to spend more time on science instruction.
The researchers based their conclusions on information gathered from state education and assessment offices last April. They also drew on prior surveys of state assessment procedures conducted by the Education Commission of the States and the Congressional Office of Technology Assessment.
The E.T.S. study focuses on tests administered to 6th- and 7th-grade students--a cohort that is most indicative of the national standard of science assessment, the researchers argued, because curricula in those grades would be “more likely to reflect the science literacy goals than would the more discipline-oriented high-school grades.”
And, they added, middle-school science has long been identified as a “key filter,” or the point in most student’s education when they decide whether to continue taking science courses.
The E.T.S. study found that, while some states--notably California, New York, and Connecticut--have moved to include elements of performance-based assessment in their testing programs, the majority of tests are more rudimentary.
“In other words,” they wrote, “short, highly sampled, multiple-choice tests dominate.”