New Test, New Rules: What Does New Jersey’s 75% Graduation-Readiness Rate Actually Mean?

NJ set 450 as the passing score on the new adaptive high school graduation assessments – the number the class of 2027 will need to hit in order to be considered ‘graduation ready’ under the first pathway.

The State BOE resolution, approved yesterday connected to the change creates different cut points for ELA and math but both end up being converted to 450. The presentation showed 75.1% of students who took the new assessment were graduation ready in ELA which is 6.2 percentage points lower than 2025. In Math, the grad-ready rate went up 1.8 percentage points from 2025 to 60.3% under the new assessment.

Those shouldn’t automatically be read as an improvement or decline because both groups didn’t get the same test; as we’ve been writing NJ replaced the prior (fixed form) assessment (NJGPA) with an adaptive test.

With the new test, the questions a student gets depend on how they answered groups of questions earlier in the test – so different students might get different question combinations and levels of difficulty.

The ELA test also changed from two extended-writing tasks to one, places more emphasis on reading and requires students to read and process more passages than the prior fixed form test. Math no longer includes extended-response items and now relies on machine-scorable questions.

NJDOE’s own August 5 presentation identified four factors that we should keep in mind when reviewing the change in rates: The move from fixed-form to adaptive testing, alignment to new academic standards, updated performance-level descriptors and redesigned test blueprints.

The presentation also acknowledges that changes to the test can affect performance until students become used to the new way of testing.

So the comparison between 2025 scores and 2026 scores can’t be assumed to be apples to apples.

Both tests also used different scales, so the change doesn’t mean NJ lowered the graduation requirement by 275 points; the new assessment uses a scale that ranges from 300 to 650.

The more important question now is whether the State completed and publicly released enough of the technical work behind these changes to explain validity, reliability and comparability.

That question has been at the center of NJ21st’s ongoing adaptive-testing investigation.

We’ve asked the state for the information that would answer these questions along with records connecting to how the new tests comply with federal requirements under ESSA.

In response to one OPRA request the DOE responded that the records identified in that portion of the request weren’t made or maintained by the DOE or any of its contractors or vendors.

Since it looks like the process has reached a completion point we can’t categorically say that no technical records exist today without submitting a new request (one is pending).

What we can say is that the records we’ve been asking for still haven’t been released to the public.

The information that IS available lets us know who helped set the standards and how the recommendation was developed, but it’s still a high-level summary that leaves a host of critical questions unanswered.

The Board resolution says the cut-score process aligned with best practices in determining validity and cut scores but the presentation doesn’t include the study results, data analyses, reliability statistics, validity evidence or outside technical-review findings that would allow us to evaluate those claims.

The same problem pops up with comparability which is necessary when trying to figure out how districts are performing over time. One of the concerns we had connected to the trending declines in proficiency and whether the shift to adaptive testing would compromise our ability to evaluate that moving forward. The State placed the 2025 and 2026 readiness percentages next to each other but that doesn’t really tell us much-their own presentation explains why the results may have changed for reasons beyond student performance.

So if no linking or equating study was done then the State should just say that and explain how the public should interpret year-to-year changes. That would be a huge issue but at least we would know.

If such a study does exist then they need to release it so we understand what these scores mean.

If you recall, we also looked at a massive 749 Cambium proposal (the State’s testing contractor) which showed us the work that needed to be done but it’s not proof that it actually was done or that the finished product produces results that can be directly compared with the previous assessment.

According to the DOE, individual student reports (digital), district summaries and aggregate school /district reports will be available to Districts on August 17. Student reports on paper are set to be sent to districts by September 4 and those need to get to parents within 30 days from when Districts get them, making October 5 the NJDOE deadline.

Districts also have to report their assessment results to BOEs and need to include aggregate results, subgroup data, comparative analyses and information about intervention strategies.

In the end, the question isn’t just about whether a 75.1% English language arts readiness rate is good or bad but what that number actually measures and whether the public has enough information to separate changes in student performance from changes in the assessment itself.

References and NJ21st Reporting