Preliminary results from the state’s redesigned junior year graduation exam show fewer students meeting the reading standard than last year, but education officials say the drop reflects an entirely new adaptive test format rather than any real decline in what students actually know.
Reading Scores Dropped on New Jersey’s New Graduation Test. Here’s Why That’s Not the Whole Story.
Preliminary results from New Jersey’s redesigned junior year graduation exam show a lower share of students meeting the state’s reading proficiency standard compared to last year, a headline number that looks concerning at first glance but comes with a significant asterisk attached. According to preliminary data for the Class of 2027, 75.1 percent of 11th graders met the reading standard this spring, while 60.3 percent met the math standard, both figures down from prior years. Education officials and testing experts are urging families and school districts not to read those numbers as evidence that students suddenly got worse at reading and math over a single school year, because the test itself changed in ways that make a direct year over year comparison statistically invalid.
What Actually Changed About the Test
The shift traces back to New Jersey’s rollout of the New Jersey Graduation Proficiency Assessment-Adaptive, or NJGPA-A, which replaced the state’s previous fixed form graduation exam starting this school year. Developed by the Department of Education in partnership with testing vendor Cambium Assessment, the new exam abandons the old model, in which every student answered an identical set of questions, in favor of a computer adaptive format that adjusts in real time based on how a student performs. A correct answer triggers a harder question. An incorrect answer triggers an easier one. The goal, according to the state, is a more precise read on what an individual student actually knows, rather than a blunt pass or fail measured against a single fixed question set that may sit well above or below where a given student is actually working.
That precision comes at a cost to year over year comparability. Because no two students necessarily see the same set of questions, and because the entire underlying question bank and difficulty calibration is new, this year’s results cannot be lined up directly against last year’s fixed form results the way earlier NJGPA administrations could be compared to each other. The state has effectively established a brand new baseline this spring, and testing experts caution that the first year or two of any new adaptive assessment typically produces exactly this kind of shift in reported proficiency rates, not because students changed, but because the yardstick did.
A Test Still Built Around the Same Standards, With Real Safety Nets
Despite the format change, the NJGPA-A continues to measure the same New Jersey Student Learning Standards for English language arts and mathematics that the previous exam covered, meaning the underlying academic expectations for graduating seniors have not shifted, only the mechanism used to measure whether students are meeting them. The state has also preserved multiple pathways for students who do not pass on their first attempt, including retesting opportunities, state approved substitute assessments, and a portfolio appeal process that lets students demonstrate proficiency through an alternative body of work rather than a single test score alone.
Those safety nets matter given how much weight the exam carries. All 11th graders, including most students with disabilities, are required to take the ELA and math components of the NJGPA-A to satisfy New Jersey’s high school graduation assessment requirement, making this the exam that ultimately determines whether a junior is considered on track to graduate. With the state establishing new cut scores, the passing thresholds that separate proficient from not proficient, specifically for this adaptive format, some shift in the percentage of students labeled graduation ready was expected as an inherent feature of the transition rather than a surprise finding buried in the data.
The Bigger Picture Nobody Should Lose Sight Of
Zoomed out to a national scale, New Jersey’s school system remains among the strongest in the country by most independent measures, a context that matters when a single preliminary data point starts generating headlines. A recent WalletHub analysis ranked New Jersey’s school system third best in the nation, trailing only Massachusetts and Connecticut, driven in large part by the state’s second best reading scores nationally for fourth and eighth graders and a high school dropout rate tied with Mississippi for the lowest in the country. Those rankings reflect a broader, more established body of assessment data built up over years, not a single new test administered for the first time this spring, and they offer useful context for families trying to weigh how seriously to take one preliminary number from a brand new exam.
None of that context erases the real work ahead for districts adjusting to a new testing format, training staff, recalibrating expectations, and helping students and parents understand what a given score actually means under the new system. But education officials have been fairly direct in their message this month, that a lower headline proficiency rate on a redesigned test says more about the test itself in its debut year than it does about any sudden change in how well New Jersey’s 11th graders can read or do math, and that families should wait for a fuller, multi year data set before drawing sweeping conclusions from a single preliminary release.
