New evidence shows that the Association of Social Work Boards (ASWB) misinformed examinees about recent exam format changes. They then quietly updated their exam handbook two weeks after the changes had taken effect. Even if examinees had been correctly informed, the format changes appear to violate testing industry standards.
What changed in the ASWB exams
On March 31 of this year, ASWB switched to PearsonVUE for the delivery of its exams. PearsonVUE is one of a handful of companies that specializes in exam delivery for a wide variety of public and private entities. It operates testing centers across the country.
With the change in administrators came a change in exam format. Rather than the exam being delivered in a single, four-hour block of time, ASWB’s exams are now delivered in two 85-question halves, each with a two-hour time limit. Examinees get a structured break of up to 10 minutes between the two halves. They cannot return to questions from the first half of the test once that half is completed.
Many examinees likely welcomed the addition of a structured break. But it probably isn’t helpful to everyone, and in some cases is probably harmful. Those examinees with test anxiety, or those who struggle with time management, now have to face a clock ticking down to zero twice in their exam. Any time left when the examinee completes one half of the exam now cannot be used to review responses for the other half, impacting how some examinees strategize time use during their test.
These impacts are exactly why, any time there is going to be a change in the structure of an exam, examinees have to be given accurate and adequate notice, and an opportunity to prepare for the revised format. Examinees can tell you this firsthand.
Exam developers also must make sure that the new format is equivalent to the old one from a measurement perspective. At worst, changes in exam format can introduce new sources of bias in exam results. So developers must make sure that any impacts on performance due to a format change are known and accounted for, and don’t impact different populations differently. These obligations are part of the testing-industry standards that, when followed, support the legal defensibility of any exam process for professional licensure. Those standards do not appear to have been followed in this case.
What examinees were initially told
In February, ASWB published a version of their Exam Guidebook for those who would be testing on or after March 31 through PearsonVUE. The guidebook provides examinees detailed structure, policy, and content information about their upcoming exams. The version of the ASWB guidebook released in February for March-31-and-later examinees included the following (pay particular attention to the “Scheduled break” section):
Note the differences between what examinees were told in this version of the guidebook, and what the actual changes to the exam were.
- Examinees were told “You will be given the entire exam time [four hours] at the beginning of the test.” They actually were only given half of it.
- Examinees were told that they had “the option of taking a scheduled break.” This was described as one of two types of breaks that you “may” — not must — take. While it could be as long or short as the examinee wished, the existence of a break between sections was actually required. Even if the examinee preferred to take the test in a continuous four-hour block of time, they could not.
- They were told that if they took that break, “the clock will stop,” suggesting that if they used less than two hours for the first half of the test, that remaining time would be available to them in the second half. It wasn’t.
This version of the guidebook was uploaded to the URL https://www.aswb.org/wp-content/uploads/2025/02/2025-ASWB-Examination-Guidebook-Pearson-VUE.pdf (emphasis mine). That URL aligns with the February 2025 release of this guidebook version.
For those who would be testing in February or early March through PSI instead of PearsonVUE, the older version of the guidebook (under the URL https://www.aswb.org/wp-content/uploads/2023/10/ASWB-Exam-Guidebook.pdf , suggesting an October 2023 upload) was still available on the ASWB guidebook page. Here’s what that page looked like in February and early March, with links to both the PSI (through March 16) and PearsonVUE (as of March 31) versions.
The October 2023 version of the guidebook, as it was available in February, appears to have accurately described the exam process for those who tested through PSI on or before March 16, 2025. Those examinees were told that the exam was delivered in a single, four-hour block of time. They could take breaks if they wished, but the clock would continue to run, and the full length of the test would remain accessible to them to review and change answers.
What changed in the guidebook — and when
On April 10, almost two weeks into tests being delivered with the new exam format, ASWB published a blog post announcing the addition of the scheduled break. While this blog post appears to be consistent with what examinees were actually experiencing, it was not consistent with what examinees had been told up to that time in the guidebook.
Four days later, on April 14, ASWB published an updated version of the guidebook for those testing with PearsonVUE — and placed it at the 2025/02/ URL. These revisions align with what examinees appear to have experienced since the start of testing at PearsonVUE: Two 85-question halves, each with its own two-hour time limit. See what this version says under “Scheduled break,” and compare that with the prior version:
Those are significant changes. Turns out you’re not given your full time at the beginning, submitting the first section of questions doesn’t “stop the clock” allowing you to save any extra time for the second section, and if you don’t want that structured break at the two-hour mark, tough luck — once two hours are up, that section gets submitted no matter what.
It is odd, to say the least, that ASWB published these revisions almost two weeks after the testing process had already been changed.
That same day, ASWB also revised the older PSI guidebook — the one located at the October 2023 URL. The updated version of that file now also describes an exam delivered in two 85-question halves, which is decidedly not what examinees experienced from October 2023 through March 16, 2025 at PSI testing locations.
Awaiting an explanation from ASWB
I reached out to ASWB for comment on Friday, asking specific questions about the exam process changes and the sequence of changes to their guidebook. As of the time of publication, I had not received a response. I’ll update this post if they respond.
In the meantime, I can only speak to the potential impacts of the revisions. Revising the guidebook in this way, and reusing the old URLs rather than using updated ones, has the same impact as backdating a signature: It gives the appearance that you’re trying to pass something off as having happened earlier than it actually did. In my opinion, this has the look of ASWB having botched the rollout of a hastily-conceived new exam format, and then trying ineptly to cover its tracks. Even today, the ASWB guidebook page on its web site notes that the April 14 version of the guidebook should be considered the most current, but is not specific about what was changed at that time — it just says that it “clarifies information” about the break policy.
Regardless of intent, the reuse of old URLs makes it more difficult for licensing boards and other stakeholders to get a clear understanding of what changed in the testing process, and when. Test developers should be continually working toward greater transparency, even at the cost of their own convenience.
Even if the guidebook had been accurate and consistent in its description of the format changes, those changes would still be problematic. No publicly-available evidence suggests the new format was checked for measurement equivalency or possible disparate impacts prior to implementation, as testing industry standards require. As I see it, the odd rollout and shifting descriptions only make ASWB look less competent.
Examinees and licensing boards deserve a more complete explanation of what happened — and repair for the harms caused. They deserve a better, fairer, more thoroughly researched exam process.
Reminders of AERA standards
As should be obvious, the developers of high-stakes exams have to give examinees accurate information. When they fail to do so, they undermine not just individual examinees, but the entire testing process.
The defensibility of any exam process for professional licensure rests significantly on that testing process abiding by established industry standards. Those standards come from the AERA Standards for Educational and Psychological Testing. (The standards are open-access; you can read them in full.)
I’ve previously documented in peer-reviewed research that ASWB’s clinical exam, like other clinical exams in mental health care, doesn’t actually follow those AERA standards in a variety of important ways. But the implementation of this new exam format, especially in a manner that was inconsistent with what examinees were told, appears to have violated several additional standards:
- Standard 4.4: “If test developers prepare different versions of a test with some change to the test specifications, they should document the content and psychometric specifications of each version. The documentation should describe the impact of differences among versions on the validity of score interpretations for intended uses and on the precision and comparability of scores.” This standard refers to measurement equivalency — the idea that changing versions or structures of a test shouldn’t harm examinees. There’s no public evidence, to my knowledge, that ASWB bothered to check whether the change in format would result in scores improving, falling, or otherwise changing. They also don’t appear to have bothered to check whether the change would impact different populations differently.
- Standard 6.5: “Test takers should be provided appropriate instructions, practice, and other support necessary to reduce construct-irrelevant variance.” As noted earlier, this kind of change in exam structure doesn’t impact everyone the same way. Such differences may be magnified when examinees are not given adequate time to prepare for changes in exam format. Many examinees spend months preparing for their tests and taking practice exams, often at great expense.
- Standard 7.8: “Test documentation should include detailed instructions on how a test is to be administered and scored.” The AERA standards, in commenting on this requirement, specifically note that the detailed instructions must include instructions on exam timing.
This all just adds to the evidence that clinical exams for mental health licensure in general — and ASWB’s in particular — do not deserve the trust of licensing boards, policymakers, or the public. At the very least, states should suspend their use of AWSB’s exams.
Edited May 19 – The “In the meantime” paragraph initially concluded by saying that the current ASWB guidebook web page gives no indication of what was changed on April 14. This sentence was edited to include that the page reports having “clarified” the break policy, though the page does not report what those clarifications were.
Reinforcing the statements in the footer, the opinions expressed here are those of the author. Nothing here should be read as legal advice or conclusions; for those, consult a qualified attorney. Thanks to Professor Matthew DeCarlo for his contributions to this article. Check out his related piece at Open Social Work.