By Jay Mathews
Washington Post Staff Writer
Wednesday, September 30, 1998; Page A3

The nation's largest study examining the use of computers in schools has concluded that the $5 billion being spent each year on educational technology is actually hurting children in many cases because the computers aren't being put to good use.

The study of nearly 14,000 fourth and eighth graders released yesterday showed that students who spent more time on computers in school actually scored worse on math tests than students who spent less time with computers. The students' lower scores appeared to be caused by the ineffective but widespread use of computers for repetitive math drills, instead of simulations and real-life applications of math concepts, computer uses that seem to improve math scores.

The research by New Jersey-based Educational Testing Service offers the first solid evidence of what works and what doesn't when computers are used in the nation's classrooms. While educators have known for some time that having PCs in the classroom was useful in teaching computer use itself, several earlier studies have provided no conclusive evidence that the new technology was any better as a tool than pencils and paper in teaching children to read, write and do mathematics.

Overall, the new study found, computers can be an important learning tool, but only in certain circumstances and when teachers are well-skilled in their use.

The study's author, Harold Wenglinsky, also identified a troubling racial element in the varying use of computers. His work indicated that black children use computers to learn mathematics somewhat more often than white children, but that blacks are far more likely than whites to engage in the less useful drill and practice exercises.

The new study "is a step in the right direction," but "I wish we did not have to have tens of billions of dollars go down the drain to reach this point," said William L. Rukeyser, whose Woodland, Calif.-based organization Learning in the Real World is devoted to the study of educational technology.

The research was paid for by ETS and published by the Bethesda-based newspaper Education Week with funding from the Milken Family Foundation. The foundation's best-known executive, financier Michael Milken, has been investing heavily in new technology for classrooms.

Wenglinsky used the test scores from a nationally representative sample of about 6,600 fourth graders and 7,100 eighth graders who took the mathematics section of the 1996 National Assessment of Educational Progress, the most extensive study of student achievement in the country. In 1996 the test for the first time asked teachers how they used computers, allowing Wenglinsky to focus on how different uses affected test scores.

Wenglinsky said that after applying statistical techniques to eliminate the influence of factors such as family income, class size and teacher qualifications, he found a strong connection between certain kinds of technology use, higher scores and improved school climate.

Eighth graders whose teachers used computers for simulations of such concepts as velocity, by showing the up and down movements of an elevator alongside a graph of its changing speed, for example, scored higher by two-fifths of a grade level than students using computers in other ways. Eighth graders who used computers primarily for drill and practice – for such exercises as dividing fractions, for example – scored more than half a grade lower than students who used the computers in other ways.

MaryJo Watson, an instructional technology specialist for the Fairfax County schools, said the study's negative view of drill and practice fits her observations. "When you take the same material that was on paper, there is not much more to it when you put it on the computer," she said. "It still does not engage the student."

The research showed no discernible difference in scores of fourth graders whose teachers used simulations and applications. Wenglinsky suggested this might be due to the scarcity of teachers who use those methods with such young students.

Fourth graders whose teachers had them do mathematical learning games on computers, however, scored about 15 percent of a grade level better than those who did not.

Perhaps not surprisingly, students whose teachers had been trained in teaching with computers did better than students whose teachers lacked such training. Low-income and black students were the least likely to have teachers who exploited all of the computers' possible uses.

In an interview, Wenglinsky said he could not reach definite conclusions about the impact of other kinds of computer use. The fact that students who use computers more in school have lower scores may reflect that "above a certain level of use they are doing things that are more entertainment-oriented."