The cut-off is a score, as I understand it, that is where only 1% of the kids that took the test in that state are at that score or above it. So, no, the score isn’t adjusted for “representative of expected scores with full participation”. It’s just simply placed where only 1% of the kids that took the test are above it.
So, if only 1000 kids in a state take the test, then 10 will be at or above the cut-off.
If 10,000 kids in a state take the test, then 100 will be at or above the cut-off.
I wonder if the schools that tend to administer in the first half of October are quantifiably different than the schools that tend to administer in the second half of October? Perhaps the averages in some states look high because the early schools are more “college prep” and the later schools are more “everyone can attend”? (And maybe the opposite holds true in states that seem to have a low average right now)
Probably just waiting until the other test results are collated and averaged and reported (in 10 days or so) is going to answer all of these wonderings…
That is incorrect. The 1% is a rough number for ease of speak. There are approximately 16000 NMSF, proportionally allocated to states not by participation, but by size of graduating class. In practice this means in some states you can make it even if you are bellow the top 1% while in states like NJ you have to be in the top 0.5%
This is an old report, but NMSC has not changed the numbers in a long time.
Data points to lower averages where there is high participation, because low scores offset the high ones. Where the PSAT is less popular, only the academically stronger students bother with the test.
Huh! So, they start with a rough “number of students we want from each state” based on population numbers. Then they set the cutoff where they would get that number of students from that state (no matter how few or many took the test). And they probably adjust the cut-off for “alternative ways to qualify” to make sure they still only get as many students as they want from each state, which is why they don’t send letters until later/next year. Interesting.
I wonder if part of the issue is that we’re dealing with means, but medians might be a more robust measure?
Also, since the number of NMSF slots is determined by the relative size of each state’s graduating class, it makes sense that a state where a higher proportion of juniors take the PSAT (e.g., New Jersey) would have a higher NMSF cutoff than a state where a low proportion of high school juniors take it (e.g., Alaska, where about half take the test). Not all potential high scoring Alaskan students take the PSAT but (effectively) all potential high scoring New Jersey students take the PSAT—and if NMSF slots were distributed by the number of test takers that wouldn’t make a difference, but the way they are distributed means that it does.
It’s not really an entirely fair system, but then again, neither would allotting slots by number of test takers. I’m not sure there’s a good answer aside from the National Merit Scholarship Corporation no longer offloading the gatekeeping step to the College Board.
I learned something new too! Sounds like that’s what TonyGrace is saying.
So how do they adjust for alternate entry? How do you fairly compare SAT to PSAT? Much easier to get a high score on SAT than PSAT (at least for my kids). For math scores, my 25D has a 800 & 36 already with zero studying, but her PSAT math came in at 740 (ouch!)
My Ds SI from her SAT would be 227. It’s not really fair IMO, but I think other than that one year, the number of AE is not that significant and you do have to have some legitimate reason for submitting.
Compass Prep does bring up the point that there were issues with the system crashing and that could lead to AEs, but in theory, the date flexibility of the digital test should reduce the number of AEs.
Also, I’m not convinced that NMSC would handle extraordinary AEs the same way… there was a lot of criticism because kids with legitimate scores got squeezed out.
If you were unable to take the PSAT you can request to qualify via AE, which means submitting an SAT instead.
The SI is calculated the exact same way, but because the SAT is out of 1600, it’s much easier to get a 1500 than it is on the PSAT. Additionally, you have the opportunity to take the SAT multiple times and you have until the June SAT, effectively an extra year of academic maturity, to take the test (you have to apply by April).
AK = 1054
CT = 952
NJ = 959
MO = 1104
TN = 1073
WA = 1024
IL= 948
Kiddo will get his November 4th SAT scores back on the 17th, it appears. He felt good about it, but said he definitely didn’t get all the math correct like he did on the PSAT. For the semifinalist qualifying score, do they have to hit above the qualifying state semifinalist score? So if the index is 217 for TN, does he have to hit at least a 217 index on his SAT to qualify? I read through the booklet but I am still confused.
…and that issue even follows you past college to grad school applications. My daughter was genuinely annoyed that she had to go back re-studying high school math to prep for the GRE.
And, she was dismayed that (because of that) she scored quite well, but not nearly as high on the GRE as she did in all prior standardized tests.
That isn’t surprising as GREs is a smaller group of higher performers than anything at HS level. I would never think a 75% ptle on SATs wouldn’t translate to a 75% ptle on GREs as underlying population of test takers is quite different and I can’t imagine many takers didn’t have to brush up on HS math.
OT: I’m still salty that they got rid of the “analogical reasoning” section of the GRE.
I mean, yeah, sure, only the linguists and the philosophers cared—most disciplines even explicitly ignored that subscore when calculating the total, in fact—but dangit, not only did we like it, we really liked it!
Did anyone get a 740 or 750 in English? What does your subject target graph bar look like? With a 740 we have the last section of one of the bars blank… What do we think? One question missed? Two? Three? I am pretty sure three would have been below a 740.