How the LSAT-Flex Has Changed the Admissions Cycle

Hi, this is Mike Spivey of the Spivey Consulting Group. It is Friday, November 20th — still relatively early in the cycle, in what's been a noticeably slow cycle. And I wanted to touch on that, particularly vis a vis the LSAT Flex, because I think that test has contributed more to the slowness than any other possible factors combined.

So let me get right to it. I want to make sure this podcast is value-added for applicants. So in other words, there's probably a lot I could say about the LSAT Flex, but if it doesn't help you as an applicant, then it's probably not worth saying. Transparency, of course, is always good. LSAC, last week, sent a number of people, including schools, an interpretive guide to the LSAT Flex, including the scale. So we've known all along — you have; I mean, you can look at the volume summaries — that scores are up at every bandwidth, but they're extremely up at the top. And this is sort of a bubble that we've never seen before. Obviously LSAC knows thisn — they have more data than than anyone — and law schools know this, and so law schools, I think, wanted to go slowly, because early in the cycle — September — they set target medians. Generally, that's like a +1, for most schools. I mean, depending where you are — at Harvard, for most cycles, it might not be a +1 at all, it might be a +0, just stay at 173. If a school is on the rise, and they're at 158, maybe it's a +2, so maybe their target median is a 160. So why has this thrown everything amok? It's because, once you start admitting, and your target's a +1, and all sudden you see this deluge of data that shows you could probably get a +2 or a +3, this cycle — every admit you make with +1 knocks you off the ability, a little bit, to get the +2 or +3. So, what I think has happened for a lot of schools, particularly schools that are data centric/data minded, is that the more they saw this LSAT data — you know, in the upper bandwidths high; you know, the percentile increase is exceptionally high — the slower they wanted to go, because they wanted to be able to pivot and change their targets.

So, LSAC released the scale bands to schools, and it was even, I think, more compelling that the LSAT Flex scores are indeed higher than we suspected relative to LSAT tests. LSAC released six hypotheses — five hypotheses — by my mind, they're missing the most important one, which is going to be mine. I won't call it the sixth because that would be in their language, so I'll call it the most likely hypothesis, and then I'll talk about the second most likely hypothesis, and then I'll talk about how this all impacts you.

Their first hypothesis is that we may be seeing a shift in when candidates are taking, within the testing cycle. In other words, higher skilled test takers, who previously might have disproportionately tested in the fall, may have opted to take the LSAT Flex in May through August.

Number two, test takers may have different levels of anxiety during test administrations, and some test takers may be less anxious testing at home.

Number three, test takers may have had more preparation time than usual for the LSAT Flex administration, since the COVID-19 pandemic resulted in the cancellation of many activities.

Number four, test takers appear to be waiting longer between test administrations to retest, so they may be taking more time to prepare than they had in the past.

Number five, candidates' perceptions of application timing may have shifted, so the customary patterns of first-time and repeat test takers may also be shifting.

So let me also say that these five hypotheses went out to every pre-law advisor, or I probably wouldn't be saying them — but your pre-law advisors have them; some of you have already heard them. The most likely of these five is number two, but I would state it in a different way. The way LSAC states it is: Test takers may have different levels of anxiety during test administrations, and some test takers may be less anxious testing at home. I would say it in the form of "state-dependent learning." You know, I'll never forget when I learned about state-dependent learning, which was in an undergraduate psychology class. The professor was talking about different cues — if you study with different cues, and then you're around those cues while you take a test, you're apt to do better. The example he gave was, if you have a beer while you're studying, you should have a beer while you're taking the test. Of course, you know, we were 19 or 20 or whatever, that was our takeaway was like, "Oh, you should drink beer, and then take tests." The real point is, if you're studying in your home office, or your home desk, or whatever, bedroom, and you're taking the test in the same exact study environment, there are reasons to believe psychologically, for a long time that your brain is more apt to retain/recall information in that surrounding.

This is nothing new, and it's probably not new to LSAC either. There's some evidence, and this is also incomplete — and everything I say right now is incomplete because the cycle data is not complete; there are still multiple LSAT Flexes to go — but, there's some evidence that bar [passage] percentage rates are higher from people who have taken online at home. I mean, you can't causate anything, but you know, there's a correlation here going on.

But even if this hypothesis is true, it still means the test is easier. Which for some reason, LSAC is just — they are sticking to their guns that the test is the same. But that wouldn't be the case if their second hypothesis is true — it should have been scaled differently then.

And I'm not saying they're necessarily wrong, but the more likely hypothesis by our estimation — and I am not in any means stating this as an absolute — in many, many, many, many years of following the LSAT, we have never seen anything like the number of test takers at the top. And if it were to start having come down, it already would have done so. So the most likely hypothesis is that this is not the same test as the former LSAT. Let that sink in for a second. There's lots of ramifications. Again, how does it impact you as an applicant? Well, on the school side, I think the schools are realizing that that is a likely hypothesis too, and if that hypothesis is true, yes, numbers are probably going to come down — I mean, of course they'll come down some, because if you score a 177, you're not going to retake; if you score 174, you're probably not going to retake, and in fact, you shouldn't. So, the high scores start trailing off over time. That happens every cycle, and LSAC has been talking about that as why the scores are going to organically come down.

But, if it is a different test, and an easier test, to begin with, you have a reserve of 4000 people who have taken it but haven't done the writing sample — so add some of those to the numbers. You have higher registrants in the upcoming tests than we had last year. And, you have retake registrants, and retakers tend to do better. So, if you score to 171, and you're retaking, it's more likely you'll score higher than lower. Most of the evidence we see points to yes, the percentiles at the top will come down, but at the end of the cycle, you're going to still have this bubble of inflation at the top, unlike anything we've seen.

LSAC, to their credit, disagrees; their psychometricians seem to believe — you know, and LSAC has really, really good psychometricians, don't get me wrong; as far as how tests are put together, they know more about this than anyone, certainly know more about it than me or the people I've been talking to. Our position is just that we can look at the raw data and say no, these numbers are going to end up way up at the end of the cycle.

So how does this impact you? The cycle's been slow, because I think schools are readjusting their targets. I mean, kudos to LSAC for sending out the data to schools, because I think this is going to speed things up. I wish they hadn't send it out to all the people they did. I'm not going to get into that, but kudos for them sending it out to schools, because I think now schools can see things and say, "Alright, well, we're going to pivot now at this right inflection point." And I think we're going to see late November, all through December, lots of admits.

One thing I want you to keep in mind as applicants is — whether it's a competitive cycle, or a less competitive cycle, schools have to do a lot of admitting. I've said this a number of times, but you know, if the average applicant is applying to 7 schools, if some are applying to 8, 9, or 10 — let's say you apply to 12 schools and you get into 10, you're saying no to 9 other schools. So schools aren't admitting 500 people to fill a class of 200; they're admitting 900 or 1000 in some cases. It's dragged out. Those admits don't come all at once; they come, you know, half in 2020 and then a large part 2021 including a large part off the waitlist. There is going to be a lot of admitting going on. There is going to be a decent amount of waitlist movement, particularly because again, I think schools are going to have to be retargeting and retargeting all cycle long, more so than ever before.

As far as test-taking, what does this mean to you? Well, if you were to believe LSAC, and you were to have what you think is your strongest score, then it means nothing! I would say that if you had an old LSAT score and you could register for an upcoming LSAT Flex, we would be pretty convinced that you would score higher. To be determined, right? So, LSAC could very well be right; at the end of the cycle maybe everything looks normal. But that, I hope, explains why this cycle has been slower, but why I also think, now, that you're going to start seeing a lot of admits post-Thanksgiving. Right after Thanksgiving, all through December, and then for some of the schools that really wait, all through January, you're gonna see lots of admits. This was Mike's Spivey, with the Spivey Consulting Group.