Scores on reading and mathematics achievement tests have risen substantially in elementary schools throughout the Washington area since the mid-1970s, as teaching has focused on the basic skills the tests record.

At the same time, paradoxically, test scores in senior highs have sagged or stayed unchanged.

With few exceptions, this pattern has occurred both in high-scoring, high-income suburbs such as Montgomery and Fairfax counties and traditionally low-achieving schools in poor neighborhoods in the District of Columbia. It has also taken place in many big cities and suburbs around the country, according to reports by test publishers and the federal government's National Institute of Education (NIE).

School officials see the elementary school gains as vindication for the more structured, rigorous curriculums and stiffer promotion policies introduced as part of a national "back-to-basics" trend. Some experts say it may also stem from applying recent research on effective schools, which stresses spending more time on important tasks, teaching step-by-step in orderly classrooms, and following a clear common plan rather the flexibility and decentralization that were the watchwords of the reformist 1960s.

Why achievement has continued to fall in high schools -- whether measured by standardized tests or college entrance exams -- is a subject of sharp debate.

Proponents of the structured curriculums note that even the most recent groups of 11th and 12th graders attended elementary school before most of these programs were introduced in the last half of the 1970s. They foresee gains in the high schools over the next few years, though some say the senior highs themselves will have to make significant changes, systematically teaching analytical skills and cutting back on "soft" electives.

Critics contend that the structured programs -- often marked by long check-lists of skills and a plethora of short "mastery" exams -- may actually have caused a decline in the more complex reasoning and analytical abilities needed to do well on end-of-high school exams.

"The gains we have made are good, but there are problems with the approach," said Richard C. Anderson, director of the Center for the Study of Reading at the University of Illinois. "I'm afraid there's been fragmentation of the curriculum with children drilling on hundreds of exercise sheets for hundreds of particular skills and then not putting it all together. The students fill in the blanks and don't write enough on their own. There's been a decline in the amount of reading and they don't read as much complex material."

Jeanne Chall, a reading expert at Harvard, retorted: "How can the students reason if they can't read? . . . . In the early grades, reading is different than it is in high school. But phonic ability is necessary for further learning even if it isn't sufficient. There's no evidence that one has taken anything away from the other."

In the Washington area, standardized test scores have risen substantially since 1974-75 through sixth grade in every school system except Prince George's County, where the scores have been virtually unchanged.

For junior highs -- grades 7 through 9 -- Fairfax was the only system to report regular gains until two years ago, and several others had slight declines. Since 1980, Arlington, Alexandria, and Montgomery also have shown some gains, particularly in math, while ninth grade averages in D.C. have climbed sharply under a program that includes considerable teaching aimed directly at the standardized tests.

Although there have been some year-to-year fluctuations, test scores for senior highs have generally moved downward in Arlington, Montgomery, and Prince George's, and stayed level in Alexandria, Fairfax, and the District.

Overall last year, Arlington, Fairfax, and Montgomery scored far above the national norms, averaging at the 70th to 75th percentile. Both Alexandria and Prince George's averaged around the norm of 50 -- meaning half the youngsters tested nationwide were above that point and half below. District of Columbia schools, despite substantial gains, still averaged considerably below the norm, at the 38th percentile.

The percentiles show how children compare with the nationwide samples tested when the exams first appeared.

It is hazardous to make precise comparisons between school systems in Maryland, Virginia, and D.C. because their testing programs use three different multiple-choice exams. However, the test publishers say their nationwide samples are equivalent, making it significant if students' scores in separate systems are approximately the same or if there are large differences.

There is a further difficulty in comparing D.C. test scores with the others because the District uses an older exam whose norms are based on how children performed in 1973. Since elementary school achievement has gone up throughout the country since then, it takes more right answers on a test to reach current national norms. This makes it likely that D.C. elementary schools would get somewhat lower percentile scores on a newer exam.

However, since achievement nationwide has stayed about the same in junior highs and dropped slightly in high schools, using a newer exam would probably make no difference in the District's percentile scores for those grades, or it might raise them slightly.

Nationally, low-income children generally do worse on standardized tests than those from middle-class homes. The average scores for blacks are lower than those for Asians and whites; Hispanics rank in between.

Schools in the Washington area generally follow these patterns. However, the variations and trends suggest that what particular schools do has an impact -- although it's difficult to determine the specific combination of ingredients that makes the difference. For example, scores for Arlington County are among the area's highest, along with Fairfax and Montgomery, even though Arlington has far more poor children.

One measure of poverty is the number of children whose family income is low enough for them to receive free or reduced price lunches under uniform federal rules. In Arlington, such youngsters amounted to 21 percent of last year's enrollment, compared to just 6 percent in Fairfax and 9.5 percent in Montgomery.

Last year blacks and Hispanics comprised almost 24 percent of Arlington's students, compared to 10 percent in Fairfax and 17 percent in Montgomery.

The two school local systems that score at the national average -- Alexandria and Prince George's -- have nearly the same racial composition, about half black, 41 to 43 percent white. But their poverty rates are much different. In Alexandria, 31 percent of last year's students received free or reduced price lunches, in Prince George's just 19 percent, a proportion lower than in Arlington which had much higher test scores.

In D.C. schools, 94 percent of the students are black and 46 percent are poor enough to get free or reduced-price lunches. The District has the most blacks, the most poverty, and the lowest test scores in the area. But since the mid-1970s its elementary students have made the greatest achievement test gains -- about 20 percentile points compared to an average rise of about 10 points for Alexandria, Arlington, Montgomery and Fairfax, when changes in their exams are taken into account.

The District did not begin making major gains until 1979 when its competency-based curriculum (CBC), developed under former superintendent Vincent E. Reed, went into effect. That also was the year of a 23-day teachers' strike and marked the start of a series of budget cutbacks.

"There's been a major pay-off from the program no matter what else was going on in the schools," said associate superintendent James T. Guines, who was in charge of drawing up the highly structured curriculum that tells how to teach major subjects in step-by-step detail.

"Yes, we have narrowed down the curriculum and we've focused on teaching those things that are valued and measured" on the tests, Guines continued. "We've been accused of teaching to the test, and we do. It's the nature of knowledge that you can't learn it all. The schools have to select a certain body of knowledge. If you do well on the tests, everybody believes you are successful. So why not teach to the test?"

Officials in the other jurisdictions said their schools also adopted systematic programs to boost achievement test scores--with detailed objectives and teaching methods, though not as tight or comprehensive as Washington's CBC.

Anderson, of the Illinois reading study center, said preparing for the standardized exams has become "an important part of the curriculum" in many school systems during the past decade, partly because of the publicity that test results now receive. Until the late 1960s, few school districts released them, Anderson noted, but public pressures for accountablity forced them out.

"If the children don't do well on those tests, the teacher looks bad, the principal looks bad, and the superintendent has egg on his face," Anderson said. "Of course they spend lots of time getting ready. That may be fine as long as you're sure that your goals are reached if the children pass the tests. There's reason to worry if you think we should be doing more."

But Jeffrey Schiller, head of testing and evaluation at the National Institute of Education, said the emphasis on standardized tests has helped elementary schools clarify their main objectives and thus improve important skills.

"There's nothing wrong with teaching to the test as long as you are not teaching particular test items," Schiller said. "Rationally, any school system should select a test that covers what they do.

"Over the past few years," Schiller added, "I think we've seen that if you clarify and narrow objectives and spend more time teaching for them, you get results. Now we have to move on to teach writing and the higher-order analytical skills."

In most European countries and Japan, preparing for national graduation exams is the prime focus of the high school curriculum. The exams themselves are much more extensive than the standardized tests in the United States.

Although the District's combination of rising test scores and budget cuts is probably unusual, the area-wide pattern of test results and school spending shows no apparent relationship between scores and spending.

For example, Alexandria schools spent $3,744 per pupil in 1980-81, compared to just $2,603 in Prince George's, but both scored at the national norm. Montgomery, Arlington and Fairfax had almost equally high scores, but Arlington, expending $3,988 per pupil, spent nearly 50 percent more than Fairfax's $2,690 and nearly 20 percent more than Montgomery's $3,392. Although the District has the area's lowest average scores, its spending per pupil -- $3,259 -- was over $500 more than either Prince George's or Fairfax's.

Across the country many studies have found the same lack of relationship between school achievement and spending or the things money usually buys, such as smaller classes. But other school characteristics, such as more time spent on learning, orderliness, and following a well-structured sequence of lessons do have a positive effect, said Michael Cohen, a senior associate at NIE.

"Our best sense is that many of these things are happening much more now than they did several years ago, especially in elementary schools," Cohen said, "and achievement is rising . . . The changes haven't moved up to the high schools yet, and the problems there are different. But I think attention is turning that way. If the same effective teaching practices are used, you would expect the skills they teach to go up too."