National Merit Cutoff Predictions Class of 2017

So if the high school performs like the Nation, 830 kids * .04 = 33 kids, but FL and Boca Rotan might perform a little better than the average???

@micgeaux – you can see the Florida report here that lists communities & schools along with the NMSFs http://www.fldoe.org/core/fileparse.php/5306/urlt/16FLSemifinalNatlMeritProgram.pdf

@micgeaux, the BIG difference this year is the three part scoring system and I think the Reading component is going to have the biggest impact as many of the highest scoring SI states have traditionally been very “mathy” - I don’t think it will be nearly as impactful for states in the middle of the index.

Case in point, my daughter scored 1440/219 due to her Reading, Writing weighting. The potential range for her PSAT score was 212 to 220 - how many other verbal kids benefited from this change in weighting as 219 could also be a 1470 PSAT score due to higher math/lower verbal scores which at 30 points higher in a significant increase?

Thanks for the information!

I was just trying to figure out if the “40 kids vs. 20 kids” held water. I’ve accounted for 33-34 kids out of the 40 with the total score of 1370 theory.

On another note, the NY data seems to make sense. If more kids scored in the 99th percentile, than 1%, the mean for NY would be skewed, unless you had a disproportionate number of kids scoring really low. Right?? I still believe in the tables. I think they will be more iffy around the 205. There is not much headroom between the 97th percentile at 200 and the 99th percentile at 205.

Compared to the 2014 SIs, the number of 99th percentiles dropped from 11 to 9; The number of 98th percentiles dropped from 7 to 3. And the number of 97th percentiles dropped from 4 to 2.

@micgeaux, based on projected concordance table completed on CC a 1370 could be as low as a 198 or as high as 212 (wow what a range!), depending on math/verbal mix, which would fall well below NMSF in NY so not certain what we’ve learned by hearing the numbers of kids who scored same.

I don’t think any comparison can be made to 2014 for SI comparison purposes as score mix is completely different


Chembiodad, nothing has changed in the scoring system for the SI. It has been 1/3 reading, 1/3 writing, and 1/3 math since 1997.

Reading has been “double” weighted now for quite some time, at least on the PSAT. The difference this year is that the regular scoring, your 1370 numbers are different. Those have a more balanced math/verbal weighting. But prior to this year, the PSAT was still 2/3 verbal, with 1/3 critical reading, 1/3 writing, and 1/3 math. The only real change there is that math only kids might be looked upon more favorably in the regular SAT total.

While it was always fairly obvious that math was only 1/3 of the PSAT, it’s more in your face so to speak this year, with the 1370 for the math kid counting less on the selection index than the 1370 verbal kid.

@PAMom21, yes, thanks for clarifying as that is my point, I think the change in the weighting in the Selection Index will both change the shape of the curve and affect individual state scoring cutoffs in mathy states

Actually, I’m not sure we are saying the same thing @Chembiodad. The weighting of the selection index has not changed from previous years.

@DoyleB, got it as I was taking the information from Test Masters write-up on 1/8/2016 so maybe I read it wrong. In my DD’s case, she went from a 215 in 2014 to a 219 in 2105 with no change in Math subscore so I think National SI Verbal %'s are going to be the driver - just my hunch.

It usually is! A math only kid will rarely make NMS. In the past, you needed to score solidly in at least two subject areas to hit most state cuts. Not only that, but the math was too easy for these math kids to show their abilities. Many of your 80 level scoring kids could have hit that score years prior to their junior year. So you ended up with quite a few 80’s, and a very harsh math curve.

@PAMom21, my twin DD’s came out of the PSAT saying math was easier this time, but as score for at least one DD was higher (no, haven’t seen other one yet) I suspect that just knew more one year later. So with DD1, we wait to see how 219 does in NJ - probably a point or two short.

Fingers crossed for you @Chembiodad! That’s right in the questionable range, so you’ll just have to wait and see. I do wish that both math and verbal had challenging questions that helped discern between those top students, but I guess that’s hard to do while also making it an SAT preparation test.

@PAMom21, agree as my DD1 scored 36 Reading, 35 English, 35 Science, and 36 Writing on ACT so more challenging questions would help cull the top of the subscore curves

<<all this="" information="" tells="" me="" that="" there="" will="" be="" relatively="" more="" nmf’s="" from="" the="" “low=”" scoring"="" states="" -="" which="" giving="" up="" as="" a="" result="" beats="" heck="" out="" of="" me.="">>

  • Sorry - that was supposed to be "low scoring SCHOOLS" and "which SCHOOL" - not states. Mea culpa. Not sure why it came out the way it did - maybe I should lay off the wine while reading this forum LOL.

Again, from the other thread:

“For example, my daughter’s School which usually has about 13 to 15 children as NMSF every year, this year she told me one of her friends already found out that almost 14 (he has not even covered the entire school yet - 1/4 of the school) has scored 218 or better including few perfect scores.”

According to the College Board SI tables, 218 is well into the 99+ range. That table smells funny. Still waiting for someone to post that kids at their school had low scores this year


@DoyleB, well that would be incredible if 1/4 were above 99%.

Given that the CB posted preliminary tables and given that others such as Test Masters have run preliminary correlations to the 2014 actuals and haven’t come up with a curve that looks like anything we are hearing on CC (it would look like an upside down hockey stick) I am going to now wait until we see what CB provided to or will provide to the the GC’s


I’m not enjoying being Captain Negative here. However, I also don’t enjoy seeing disappointment due to questionable corporate behavior. In Jed Applerouth’s blog post, he suggests possible explanations for inflated percentiles. The last bullet is:

“Something else entirely is going on.”

He clearly doesn’t want to go down that path. He has no smoking gun. Perhaps he has a vested interest in staying in the good graces of the College Board. Access to early version of their tests and information about changes to their tests is important to his business - I understand that. I’m under no such limitations.

Last year the cutoff in my state was a 218, which is in the middle of the 99 percentile bands. An equivalent point on this year’s table is a 208. There are a number of students at the school with scores in that neighborhood running around thinking they might be NMSFs. Based on the other scores I’ve heard of, and my knowledge of their academic abilities compared to previous NMSF that I know, I believe they have no chance. Zero, zip, nada, no way.

At the same time, I’ve talked to no less than 4 parents who have told me, “Junior did much better than we expected on the PSAT. We were only planning on taking the ACT, but we have signed him up for the SAT in March”.

As I’ve said previously, computing valid percentiles for the SI chart is trivial. I believe the current chart is bogus. I believe the College Board is getting their butts kicked by the ACT. I believe that inflating percentiles leads to more SAT test takers.

The above rant, of course, is only my opinion. Which is worth exactly what you paid for it, so
 At any rate, I hope I’m full of crap, completely wrong, and all of those kids become NMSF. But I doubt it.

I like this latest conspiracy theory after the series of ones like CB releasing scores early to paying customers of colleges and TASPs while failing to deliver scores to students.
They inflated percentiles to attract students to take new SAT. Lovely!

To dispel this conspiracy theory, the College Board could spend 5 minutes generating an SI percentile table based on the students who actually took the PSAT last year, and publish it. If it looks like the one they published last week, I will be the first to apologize.