Last spring, third graders across New York sat down for the state English exam in front of a screen, the first year the state required it of children that young. Their results arrived in August on the same score reports and the same four performance levels that earlier classes had earned with pencils. In New York City, the share of third graders scoring proficient in reading fell by 14 percentage points, a year after it had risen by 13.

Now more than half of the New York City Council has asked Education Commissioner Betty Rosa to waive the computer-based testing requirement and let the city's students in grades 3 through 8 take the exams on paper again. Council Member Eric Dinowitz, who made the request, argues that "a computer-based testing mandate creates pressures for additional screen exposure." A state Education Department spokesperson, J.P. O'Hare, told Chalkbeat, "The Department is reviewing Council Member Dinowitz's request and will have additional information in the coming weeks."

Our view is that Albany should not grant one district a paper exemption from a statewide test. Before it decides anything, the Education Department should publish the analysis it says shows the switch to screens had a negligible effect on scores. If paper comes back, it should come back on the same terms for every district, because a test is only useful to parents when every child takes it under comparable conditions.

Three Springs, Six Grades, One Screen

The transition took three years. According to the Education Department's computer-based testing page, fifth and eighth graders were required to test on computers in spring 2024, and fourth and sixth graders joined them in spring 2025. By spring 2026, all students in grades 3 through 8 were required to participate. Students with accommodations that call for paper kept that option, which means the system already runs two modes side by side for a smaller group of children.

The schedule matters because it put different grades on screens in different years. Fifth graders, in their third spring online, still posted substantial swings in reading scores in both the city and the state, which is part of why reporters and testing experts doubted the device alone explained the results. Third and seventh graders tested on computers for the first time last spring, and the two grades moved in opposite directions: seventh grade reading scores held steady while third grade scores in the city dropped sharply.

What Massachusetts Measured

The best evidence on the question comes from a state that ran the experiment cleanly. During its own transition, Massachusetts gave the same PARCC exam online and on paper in the same years, which let researchers Ben Backes and James Cowan of the American Institutes for Research compare otherwise similar students. In a 2019 study in the Economics of Education Review, they reported an online test penalty of about 0.10 standard deviations in math and 0.25 in English language arts. The gap "partially but not fully fades out in the second year of online testing," they wrote.

In their working version of the paper, the authors found the reading penalty was larger for students scoring near the bottom of the achievement distribution. Backes put the policy problem plainly in an interview with the Harvard Graduate School of Education: "The state should think about how schools might be penalized for something out of their control." The school's own summary of the study translated the gap into classroom time, describing online testers who performed as if they had five fewer months of preparation in math and 11 fewer months in English.

What New York Says It Found

New York's own results do not settle the question in either direction. Statewide, grades 3 through 8 reading proficiency fell to 48 percent from 53 percent, as we reported when the preliminary scores came out, while math rose to 57 percent from 55. State officials told Chalkbeat they had extensively studied whether the computer transition played a role and concluded the effect was negligible. That study has not been posted for parents to read, though the department did publish grades 3 through 8 test mode comparability reports for its 2017, 2018 and 2019 administrations.

The people closest to the test have been more cautious than the press office. Marianne Perie, who co-chairs New York's technical advisory committee and spoke in a personal capacity, told Chalkbeat in August, "My instinctive reaction is there may have been an issue last year that we missed." Andrew Ho, a Harvard professor who also serves on the committee, said, "We should absolutely look at the test, but we should absolutely also look at what the test is likely revealing." NWEA, the state's testing vendor, said it had "not found any anomalies through our standard due diligence process."

"Speculation is not a substitute for data." J.P. O'Hare, New York State Education Department, responding to economist Emily Oster

O'Hare offered that line in August when Oster, a Brown University economist, questioned the results. It is the right standard, and the department can meet it by releasing the mode analysis for the 2024, 2025 and 2026 administrations, broken out by grade. A parent in Rochester or Plattsburgh could then see what the numbers on a child's report measured.

Why a One-City Waiver Breaks the Comparison

New York City is the largest district in the state, and a paper option limited to its students would split the testing population along the one line that the Massachusetts research says can move scores. City results would be measured on paper while Buffalo, Yonkers and every rural district stayed on screens, and the gap between them would blend real differences in learning with differences in format. School accountability ratings, district comparisons and the statewide proficiency rate all assume that a 3 means the same thing everywhere.

The concern behind the Council letter is legitimate, and it is not only a city concern. Parents across the state have pushed districts to cut classroom screen time, a debate we took up in our op-ed on the FOCUS Act. A few hours of state testing each spring is a small part of a child's screen week. Families who want paper deserve an answer grounded in evidence, offered to all of them, rather than an exception carved out for a single district.

What Parents Can Ask Now

Parents do not have to wait for Albany to get useful information. A family can ask its school whether the child practiced on the testing platform before the spring window, how many sessions that practice involved, and whether devices, headphones and keyboards were the same ones students used in class. Parents of children with an IEP or a Section 504 plan can ask the team whether a paper-based accommodation is appropriate, since that option survived the transition for students who need it.

Families can also ask their board of education to request the state's mode analysis on the community's behalf. A request on a board's letterhead carries institutional weight that a single family's email does not. A district that sees its third grade scores fall in the first year on screens has a direct interest in knowing whether the device played a part.

The department has promised more information in the coming weeks. When it arrives, the most useful document it could attach would be the study it has already described, with every grade and every year of the transition included.