When used effectively, computer drills and tutorials can improve student performance in math and science—but the benefits of computer simulations and electronic sensors are less tangible, at least in studies that have been conducted so far, according to a research review that the National Science Foundation (NSF) plans to release this spring.

"The kind of applications that seemed to be more challenging or oriented to higher-learning objectives didn't fare as well as simple tutorials," said James A. Kulik, a research scientist from Michigan University who wrote the review, titled "Effects of Using Instructional Technology in Elementary and Secondary Schools: What Controlled Evaluation Studies Say."

"The modes of computer use that had the best record of helping kids learn had simpler approaches," Kulik said.

The report is still being reviewed by NSF, but its findings were described in a November briefing paper called "School Mathematics and Science Programs Benefit From Instructional Technology." It comes as the federal Education Department prepares to launch a $15 million research project of its own to evaluate the use of technology in schools.

For this report, Kulik reviewed 36 controlled evaluations of instructional technology in elementary and secondary schools published since 1990, as well as some earlier reviews and less formal studies. He did not include theoretical works, case studies, or studies that did not measure learning outcomes.

The 36 evaluations focused on four types of computer applications: integrated learning systems (ILS) in mathematics, computer tutorials in science, computer simulations in science, and microcomputer-based laboratories.

ILS in mathematics

ILS combine drill-and-practice questions and tutorial lessons. They require students to respond frequently, provide students with feedback on their answers, and keep detailed records on student performance.

Kulik reviewed 16 studies that evaluated students who used ILS to study mathematics. Each of the 16 studies found test scores were higher among students who were taught with the help of ILS software. The scores improved by a 0.38 standard deviation, or from the 50th to the 65th percentile.

The research suggested that schools used ILS only 15 percent to 30 percent of the recommended time, and that the systems were treated as an add-on to the curriculum rather than an intrinsic part of instruction.

Kulik said the results might have been even better if researchers focused on "model implementations" rather than typical ones.

"Most evaluation studies from the 1960s and the 1990s suggest that students benefit from ILS instruction in mathematics," Kulik wrote. "In the typical evaluation study of the 1980s, ILS instruction raised mathematics test scores by about 0.4 standard deviations. More important, in the typical evaluation study from the 1990s, ILS instruction raised mathematics test scores by about the same amount."