The District’s public schools are making modest strides closing the vast achievement gap between white students and students of color, according to test results released Thursday that show the chasm narrowing on the English language arts exam.

The achievement gap has frustrated educators, parents and students for decades, and the standardized test results for the 2017-2018 school year show the persistence of the problem: Even as the gap closed in English, it widened slightly in math, with improvements among white students outpacing gains by black and Hispanic students.

In announcing the results, city leaders celebrated improvements by students in every ward and across all races and income levels on their English and math tests.

But they warned that significant progress is still needed: Despite the overall improvement in scores, only about one-third of students met expectations on their math and English exams.

The need for improvement is especially acute among special education students, who had little growth in their scores.

Children whose second language is English had a slight dip in their math scores, but showed advancements in their English results.

“While we are pleased to see growth in the right direction, we all know there is much work to be done,” Mayor Muriel E. Bowser (D) said Thursday evening at a news conference.

Students in the traditional public school system and charter schools — which are publicly funded and privately operated — take the test.

“We need to get smarter and more sophisticated in our ability to assess students’ needs and their gaps and then properly model an intervention,” said Amanda Alexander, interim chancellor for D.C. Public Schools. “We have built systems for this in [D.C. Public Schools] and it is a matter of us implementing them.”

The computerized Partnership for Assessment of Readiness for College and Careers test — widely known by the acronym PARCC and administered in the spring — evaluates students on a five-point scale, with those who earn fours and fives meeting or exceeding expectations and considered “college and career ready.”

Students are tested each year in grades three through eight and then once in high school, as required by federal law.

Middle schools in the traditional public school system showed the most growth — a small win for Bowser, who made a lofty campaign promise in 2014 to significantly improve the city’s struggling middle schools. Charter schools also made gains in their middle school scores.

Citywide, about 33 percent of students met or exceeded expectations in English, a 2.8 percentage-point increase from last year. About 27 percent of students reached that benchmark in math, a 2.5 percentage-point increase from 2016-2017.

Results show that 82 percent of white students met expectations in English, compared with about 25 percent of black students and 32 percent of Hispanic students. The achievement gap between black and white students is slightly wider in math.

White students make up less than 15 percent of the city’s public school population.

“Everyone can go up, but we need to see African Americans and other demographics who haven’t always done as well go up at a faster rate,” said Will Perkins, a research and policy analyst at the 21st Century School Fund, an education advocacy organization.

This marks the fourth year that D.C. students have taken PARCC, an exam designed to be more challenging than previous tests administered by the District. It enables the performance of D.C. students to be compared with peers in a handful of states, including Maryland. The city has shown growth every year.

In Maryland, 41 percent of students in grades three through eight met expectations on the English test in 2017, while only one-third reached that benchmark in math.

Students in wealthier families tend to perform better on standardized tests, and D.C. and other school districts have encountered backlash for using the exam’s results to assess school quality. The District is introducing a new school ranking system in December that will rely heavily on test scores, allowing parents to compare charter and traditional public schools in one place.

D.C. Public Schools outperformed charter schools on the 2018 PARCC test. Overall, the traditional school system showed greater improvement over 2017 and had a higher percentage of students meeting or exceeding expectations on the tests.

In English, 31.5 percent of students in charter schools met expectations, compared with 35.1 percent in the traditional school system. The gap between the two sectors is smaller in math.

But when looking at the performance of students who are at-risk — defined as those who are homeless, are recipients of welfare or food stamps, or have languished in high school — the charters performed slightly better. In math, for example, 18.5 percent of at-risk students in the charter sector met expectations, compared with 13.4 percent in D.C. Public Schools.

Still, the traditional system and charters both showed progress in this group, which accounts for more than 40 percent of the student population.

“We are really thrilled that both public charter and [D.C. Public Schools] have shown growth and continue to improve,” said Scott Pearson, executive director of the D.C. Public Charter School Board.

Pearson and Alexander, the interim chancellor, highlighted schools that posted some of the biggest gains. Hardy and Jefferson middle schools, which are part of the traditional public school system, saw double-digit increases in their test scores.

At Ketcham Elementary School, which has one of the highest populations of homeless students in the city, scores on the English exam rose 12 percentage points.

Among charter schools, Cesar Chavez Public Policy-Chavez Prep had double-digit gains on its math scores. Basis D.C. recorded double-digit improvement in students meeting expectations on the English exam.

“We think that shows that Zip code does not need to be destiny and schools can perform well located anywhere in the city,” Pearson said.