This past school year, the percentage of Connecticut students in grades 3 to 8 who met the proficiency benchmark for math and English language arts on the Smarter Balanced assessment reached the highest levels since the test was implemented in 2014-15, according to new data from the state Department of Education.

Chief Performance Officer Ajit Gopalakrishnan repeatedly called the results "heartening" in a conference call last Friday, striking a more positive tone than when talking about SAT results last month.

"It's heartening to see some of these improvements, especially in mathematics, which historically has been a challenge for us," he said. Growth in math was particularly strong in the seventh grade.

Gopalakrishnan said about 231,000 students in grades 3-8 took the assessment this year, a participation rate close to 99 percent.

Compared to the 2017-18 school year, the share of students meeting benchmark increased by 1.3 percentage points for math and 0.4 percentage points for ELA. Gopalakrishnan was heartened to see these numbers were even higher for high-needs students – respectively, 2.1 and 1.8, the highest for any group.

High-needs students are those who are English language learners, have disabilities or are from low-income families. After high-needs students, Hispanic/Latino students saw the largest growth in ELA while Asian students saw the largest growth in math.

The groups that outpaced the overall statewide growth in both math and ELA are Asian, black and white students.

"There are still large gaps. They are getting a little bit better," Gopalakrishnan said, adding, "We still have a lot of work to do."

In 2018-19, 55.7 percent of students met or achieved benchmark in ELA, while that figure is 48.1 percent for math. Those figures are higher for the lower grades: 55 percent of third-graders and 52.5 percent of fourth-graders met benchmark in math.

Among local districts, the figures are higher than the state levels in both ELA and math for East Lyme, Groton, Lyme-Old Lyme, Montville, North Stonington, Preston, Salem, Stonington and Waterford, while they are lower for both in Ledyard, New London and Norwich.

[naviga:iframe title="Percent who met the ELA proficiency benchmark" width="100%" style="border: currentColor;" src="https://datawrapper.dwcdn.net/XzckY/3/" scrolling="no" height="550" frameborder="0"] [/naviga:iframe]

Groton, New London and Norwich are among Connecticut's 33 Alliance Districts, the ones the state considers the lowest-performing across a variety of metrics that go beyond test scores.

Gopalakrishnan said that of the 33, 29 saw higher math proficiency rates and 19 saw higher ELA proficiency rates than in 2015-16. He added that 21 of the 33 showed greater improvements than the state as a whole.

For this past school year in southeastern Connecticut, North Stonington had the highest share of students meeting benchmark in ELA, at 80.1 percent, while Lyme-Old Lyme was the highest for math, at 79.1 percent. From 2015-16 to 2018-19, Stonington saw the largest growth for math while Waterford saw the largest growth for ELA.

Statewide, New Canaan was the highest performing district in both ELA and math for 2018-19. School-level results for each district are scheduled to be released Monday on EdSight, the state Department of Education's data portal.

[naviga:iframe title="Percent who met the Math proficiency benchmark" width="100%" style="border: currentColor;" src="https://datawrapper.dwcdn.net/cgZ0G/5/" scrolling="no" height="550" frameborder="0"] [/naviga:iframe]

Gopalakrishnan said the state has reached out to several of the districts that saw the greatest improvements to talk about what they're doing, but that "there are no shortcuts to improving academic achievement."

Education Commissioner Miguel Cardona noted in a news release that a "strong instructional core" includes "connection with students, a strong curriculum, and effective teaching and leadership."

Gopalakrishnan said there are no plans to revise the test at this point, and that it "would not be appropriate to compare our scores to other states that are not using Smarter Balanced."

e.moser@theday.com