One year ago, ninth graders at Evans Junior High School in Northeast Washington scored nearly two years below the national grade average in the math category of the annual standardized test of basic skills.

This year they scored much better. School officials said that all Evans students had to take at least one reading course, teachers taught test-taking skills and gave students 20 new words a week that were used in their daily instruction.

Home economics teachers gave vocabulary lessons and even the gym teachers got into the act, assigning nine compositions during the year.

The result was that this year's ninth grade class matched the national norm of 9.8, equivalent of the eighth month of the ninth grade, while exceeding the ninth grade norms in language with a 10.2; reading, 10.9; and science, 10.7. The scores for last year's class were 8.0, 6.4, 6.5 and 5.9, respectively. "This year we didn't get off course. We stuck to it," said Evans principal Margaret Saxon.

Evans, a school of 410 students in a depressed and poverty-ridden section of East Capitol Street, is only one of many of the District's 187 schools that showed improvement over last year in the recently released scores of the Comprehensive Test of Basic Skills. The most significant improvement in scores city-wide occurred in schools such as Evans that traditionally have had low test scores.

Throughout the city, third graders and sixth graders exceeded national norms 11 of 12 times in reading, mathematics, language, reference skills, science, and social studies--the six areas tested. Their counterparts last year surpassed national norms only five of 12 times. Ninth and 11th graders remained below the national average, but showed the greatest improvement, raising their test scores by as much as a year and two months.

Eighth graders took the test for the first time in reading, math and language and scored below the national norm.

A far different picture appears when, instead of comparing the results of successive classes, the scores of the same group of students this year are compared with their standardized test scores from previous years. Using percentiles based on a national norm of 50, the scores of sixth graders this year showed dramatic improvement over their test scores as third graders in five of five categories. Reading scores, for example, rose from 38 as third graders to 52 as sixth graders.

But according to the scores, the improvement stops there. Scores actually worsen for older D.C. students. This year's percentiles for ninth and 11th graders, shows that ninth graders and 11th graders, respectively, improved their earlier scores in only one of five categories and actually did worse in several others.

School officials say this is because this year's sixth grade class had the benefit of starting earlier with the school system's competency-based curriculum, implemented in 1979, which demands mastery of certain grade level skills. Officials also say that the mid-year promotions plan, which the superintendent and some board members want abolished, has also resulted in a "purer" sixth grade class by failing slower students at the mid-point of the school year for the past three years.

"Learning basic skills is very much like walking. If you don't learn how to walk at the time that you should, you are going to have a peculiar kind of walk," James Guines, associate superintendent for instruction, said in referring to older students who had already lagged behind grade and skill levels by the time the tougher academic standards were introduced. "The older kids are always behind. The longer and earlier our kids have been in CBC competency-based curriculum the better they do."

To improve the grade equivalent test scores of this year's D.C. students, individual schools did a variety of things, including a wide array of test preparation techniques and materials and an emphasis on reading and vocabulary work in classes that traditionally have not had such emphasis. School personnel also gave practice exams--in some cases based on tests that are no longer used--to relax students who experienced anxiety attacks during test time, and emphasized the importance of test taking to those who before did not show up for the tests or quit halfway through them.

The two-year-old procedure for preparing District students to take the test is different from the preparation given students in surrounding jurisdictions. In Montgomery County, for example, officials said they destroy their outdated standardized tests and never allow teachers to use old tests, nor any other kind of standardized test of similar difficulty, as practice exams for their students. Some Fairfax County schools use sample questions in their "test taking packages," but said they are never questions from old tests.

In the District, however, teachers who want to give practice tests can obtain copies of old (Q Form and R Form), but not current (T form), CTBS tests in helping students to prepare. School system administrators say that only eight schools asked for copies of the old tests.

John Stewart, senior product manager of CTB/McGraw Hill, the test publishing division of McGraw Hill, said, "There is no similarity between the new and old tests. The questions and answers are different." But the factor that made the most difference in test scores, according to William M. Jefferson, principal at Anacostia Senior High, was teaching students how to read better.

"We emphasized that every teacher should teach reading and vocabulary," Jefferson said. "Regardless of the course, there is a certain vernacular and vocabulary they need to learn. If you improve reading you improve in other areas."

Anacostia High's Ninth and 11th grade students came closer to national norms this year by bettering last year's ninth grade classes in reading, math, language and science grade level scores by two months to nine months.

At Johnson Junior High School in Southeast, ninth grade student scores went from 5.8, 7.5, 7.7 and 5.6 last year in reading, math, language and science to 9.4, 9.5, 8.6 and 10.0 this year. Instructors there were told that part of their Teacher Appraisal Process (TAP) would look at how well their students did on test scores, said principal James C. Greene.

Johnson students also kept "skills notebooks" in their courses and parents were involved by having to sign the notebooks periodically to prove that they had seen how well or how poorly their children were performing, Greene said.

Van Ness Elementary had some of the worst test scores in Ward 2 last year. This year its third graders came in above the national norm in reading, math, language and science. Its sixth graders exceeded the national average in math and science.

"People at the regional school system office offered us help and asked what our problems were," said Van Ness principal James Harris. "Our teachers had a deeper sense of what they had to do." Van Ness guidance counselor Bernice Brown used the school's public address system and cassettes from the school system's guidance department to go over test taking techniques and sample questions every Friday morning.

"The children last year were uptight, nervous, not relaxed," Brown said. "The practice tests helped that and gave the students a chance for immediate reward. They were anxious to find out the right answers afterward."