Saturday, February 20, 2016

New National Merit Class of 2017 cutoff predictions: Commended, 207, Mississippi 209-210

UPDATE 2/27/2016: I have updated the headline to this post to change my prediction for Commended status to 207. It's also possible that a score of 209 in Mississippi could meet the cutoff. I misread the percentile chart on the Hispanic scores by one percent, and as a result I may have beeen one point too conservative in my estimates. With that said, Commended may still be a 208; the scores are, as I've said, incredibly clumpy.

    Several weeks ago I wrote about the College Board’s release of PSAT scores and gave my prediction for Mississippi’s cutoff score for the Class of 2017 – in other words, the “Selection Index” score required to be named a National Merit Semifinalist. I said it would be a 204. Ha!
    This entire process is a mystery to most people, but each state gets its own cutoff score, designed to recognize the top one percent of students in each state. These cutoffs vary according to how well students in a state perform, so Mississippi and West Virginia, for example, have lower cutoff scores than Massachusetts and Connecticut.
    I’ve written about the PSAT extensively for several years, and you can read some of my old blog posts by clicking here. Earning National Merit Finalist status is a pretty sweet deal, since it can result in free-ride scholarships at a number of good universities.
Just add 6 or 7
    There have been some major changes in the PSAT, and juniors this year took an entirely new test with a new scoring system. I believe the PSAT for 2015 was poorly designed, to put it mildly, in that it tended to be difficult for average students, while the brightest students were able to complete the test with few missed questions. So the test didn't provide the differentiation on the high end that was needed.
    Prior to officially administering the test, the College Board gave the new PSAT to a large representative sample of 11th graders to establish score percentile norms, but it is my opinion that they forgot to account for “preppers” when creating this sample. It has become increasingly common for very bright students to “prep” for the PSAT, and a few high schools with a relatively small number of “preppers” can completely throw the score distribution for a loop. That’s what I think happened, and why I think there was a one-month delay in the release of the scores. I think the College Board executives were just gobsmacked by the results and probably spent a month with everyone running in circles wringing their hands.
    Based on a Selection Index chart release by the College Board, I predicted back on Jan. 13 that Mississippi’s cutoff score would be 204, with the caveat that my prediction was based on the information they provided. Well, the information they provided is now believed to be pretty lousy, to put it mildly. Based on their charts, schools all over were quietly reporting that they had a bumper crop of National Merit winners. Well, no, it just ain’t so.
    On Feb. 8, based on some new information I updated my original column and wrote that Mississippi’s cutoff would be higher, and I thought Mississippi’s cutoff would be 206, 207, or perhaps even higher. Well, I’m pretty sure it’s going to be higher still, and I’m sorry for that, because I know some of the kids who are likely to be left out of the process. I’m hoping my kid won’t be one of them.
    I have been snooping on the website, and apparently the cutoff scores for the National Hispanic Recognition Program are out, roughly six months ahead of those for regular Semifinalists. The NHRP recognizes Hispanics who score in the top one percent of Hispanics taking the test in their region. By looking at these scores, we can get an idea of where other scores might fall.
    For example, last year the Hispanic cutoff score for the “South” region was 199, which was in the mid-range of the 95th percentile. This year it’s 204. Since ethnic groups tend to perform the same over time, we can be reasonably certain a score of 204 is in the 95th percentile.
    By adjusting the scores in such a way as to put 204 into the 95th percentile on the faulty SI chart, I have concluded that a score of 208 is going to be needed for Commended status; there is some chance that a score of 207 will do the trick.
    I won’t bore you with any more of my reasoning, save to say that I believe the formula for determining most sub-99 percentile state’s cutoff scores is to do the following:
    1. Study your state’s cutoff scores over the past several years and decide what percentile, including tenths, best represents your state.
    2. Use the VERY faulty Selection Index that the College Board provided to find the score which you feel most closely matches your selected percentile. I’ve included this incorrect and faulty index on this page. This is obviously an inexact science!
    3. Add seven. For states with cutoffs very close to or just inside the 99th percentile, adding six might do the trick. To be conservative, add eight. (If you add nine and still make the cut, you can start celebrating!).
    Mississippi’s cutoff score has been rising, primarily due to year-long prep classes being conducted by DeSoto County Schools and Madison Central High School, among others. Last year it was 209, which was roughly in the 98.5th percentile. The scores were supposed to drop this year due to the new test with a lowered top score, but they aren’t going to.
    If we look at this year’s incorrect College Board SI chart, we find that a score in the 98.5th percentile is roughly equal to a score of 203. If we add seven to this we get 210, which is now my prediction for Mississippi’s National Merit cutoff score. I am not particularly happy with this prediction, but it is what it is.
    Predictions are just that. The cutoff could be 209, or 211, or God forbid, even 212, at which point there will be tears at my house (from me). But 210 or 211 seem like the numbers to watch, and 208 seems most likely for Commended, with the possibility of 207 making the cut.
    Note, by the way, how incredibly compressed these scores are. The Class of 2016 had a Commended score of 202 and a Mississippi cutoff of 209, a difference of seven points. This year I'm suggesting the Commended cutoff will be 208 and the Mississippi cutoff will be 210 or 211. These percentiles are so clumpy that the National Merit Corporation may have a very difficult time setting cutoffs. A Mississippi cutoff of 209 might represent two percent of our state's students while a cutoff of 210 might only include one-half of one percent. I predict massive problems and some squawking over how this is dealt with.
    The College Board did a real disservice to test takers by putting out false percentile charts. It created a lot of false hopes, and I think they have just done a terrible job with everything they have done with the new PSAT. And it’s not just the PSAT; they’ve damaged the SAT as well, with all kinds of Three Stooges types of stupidity. It may not be long before the SAT will be featured in a business school case study on “How to Destroy a Brand.”
    But that’s another story. My prediction for Class of 2017 Commended cutoff: 208, maybe 207. Mississippi cutoff: 210, 60%; 209, 10%; 211, 20%; 212, 10%.
    I’ve made two earlier predictions which I decided were wrong. I may return next week with a different song. But I think we are slowly bracketing in on what scores will be needed for the brass ring, despite the efforts of the College Board to keep us all in the dark. They could provide all of this information for us, but they just won't.


Unknown said...

So, I guess you are saying that a student who has a score that is in the 99th percentile (1380) did not really score in the 99th percentile.

Col. Reb Sez said...


The College Board figured the percentiles differently this year. In the past, a score which was reported to be in the 99th percentile was in the 99th percentile of a sample of those who actually took the PSAT. This year the 99th percentile is based on a sample of all students nationally. The group that takes the PSAT is always a "smarter" group. It's thought that the College Board did this so people would feel better about their scores and would be more likely to take the SAT over the ACT.

The 2015 Understanding your PSAT score guide lists percentiles for both the National and User sample. A score of 1380 is in the 99th percentile of the National sample, but only the 98th percentile of the predicted Test-taker sample. The real question people have is whether the Test-taker sample was valid; the thinking is that the College Board failed to account for the increasing number of preppers at some schools (in Mississippi, DeSoto County and Madison Central, for example). If so, it could push that 98th percentile down to the 97th percentile; it's just hard to say.

I do think I may have been overly pessimistic on both the Commended and Mississippi Selection Index cutoffs. I think the Commended could be as low as 206, and that Mississippi is likely to be 209. I hate to say the latter, as I would rather err on the side of caution.

Blanche Martin said...

Col. Reb, Thank you for this information. My son is in Louisiana. He has an SI of 212. Do you have any predictions as to what the cutoff might be in Louisiana? I appreciate any information you may be able to give me.

Thank you,
Blanche Martin