What’s going on with California’s MFT Clinical Exam? [Updated]

California flagIt is certainly debatable what an ideal pass rate for licensing exams should be. If the pass rate is high, that means almost everyone gets through. Then the tests don’t serve a meaningful function. (That’s pretty much the status quo.) If the pass rate is low, it raises questions about the validity of the exam, given how much time most examinees spend preparing for it. But what makes a pass rate too high or too low? Given that the exams don’t do much of anything anyway, it’s hard to say for sure.

But it does raise eyebrows when pass rates for a single exam fall off a cliff, as seems to have happened for California MFTs over the past year. Here’s the data provided by the state’s Board of Behavioral Sciences for first-time test-takers on the state’s clinical exams:

Counselors had a rough go of it in the third quarter of 2016, but that could simply be random noise. The LPCC license is still fairly new in California, and so very few people actually take the clinical exam. There were about 20 first-time test-takers per quarter shown here. For MFTs, however, the sample size is much larger, with several hundred first-time test-takers each quarter since the new exam began in early 2016.

It is normal for tests to fluctuate a bit in difficulty with each test cycle, but on the MFT Clinical Exam, the passing score is supposed to fluctuate to account for that. Tougher exams have lower score requirements to pass. It doesn’t stand to reason that the examinee pool has changed dramatically over a single year. So how is it that the pass rate has dropped from nine out of 10 to just more than half?

I’ve reached out to the BBS for comment, and will update this post when I hear back. Notably, California is the one state in the country that doesn’t use the National MFT Exam. It uses a state-based exam instead.

Update, May 8, 2 pm: BBS Assistant Executive Officer Steve Sodergren provided this explanation from their Exam Development office (emphasis mine):

In a typical examination cycle, the candidate distribution is comprised of the following three groups: (1) first-time test takers who achieve a passing score, (2) first-time test takers who do not achieve a passing score, and (3) repeat test takers. When the LMFTCE exam went into effect in January of 2016, the candidate distribution was atypical. All candidates who sat for that examination had taken and passed the Standard Written examination, and many had also taken the Clinical Vignette exam. It appears the distribution remained atypical for the next few subsequent administrations. Due to this anomalous situation, a higher passing rate for these administrations was expected to be seen than those for a typical distribution.

That would be a sensible explanation if the pass rate for first-time test-takers in 2016Q1 were unreasonably low, but it doesn’t make a lot of sense for the pass rate to be unusually high. Repeat test-takers typically pass an exam at a lower rate than those taking an exam for the first time. As those repeat test-takers from the (now outdated) WCV became a lower and lower proportion of what the BBS data would call “first-time” examinees, the pass rate for first-time test-takers should come up, not down. I’ll gather more info and post another update if warranted.