I say all of our bottom danglers in the SI cutoffs move to Wyoming and retake the PSAT!
@Speedy2019 so we know it’s probably between 4 and 4.5 million. LOL.
So the SI percentiles are based on a research study? It figures. I wonder if it’s even “weighted” for users (as opposed to all 11th graders)? Anyone know?
@Mamelot, Mme, I believe you are serious I follow your posts in this CC thread and others - you are very knowledgeable compare to me.
Again
“Please, do not take my opinions or conclusions seriously since I am junior in HS and if you follow my posts you would see things like “my methods or my ways of estimate have many flaws” or " an estimate is an estimate. My Guess is as good as or as bad as anybody’s”. I am not a CB official (from msg 4120)"
@OHToCollege I think we are saying the same thing. If I understand correctly, commended is the top 50,000 students and is not divvied up by state. NMSF is divvied up by the number of students in the graduating class compared to the nation. (I guess.)
Exactly, but for 2000 CA kids to surpass her 217 then all of them would come from the national 99+ pool. So 2000 out of 8000 nationally or 25% of that pool would have to come from CA (which only accounts for 11-12% of the total test takers), that seems incredibly improbable. Of course if her 217 being 99+ is a fraud then all bets are off.
OK Sure- Everyone - Here is how NMS people come up with Commended and Semi Finalists (Hint: They do NOT use concordance, they do NOT use percentiles:)
They create an excel spreadsheet with 1.6 million students listed in rows with their Selection Index score listed in columnn B. Sort by SI, highest to lowest. When they hit about row number 50,000 (up or down a few rows so that all students of the same SI are covered) they look at that column B and say “Commended cutoff = That number”! Yeah. We have commended.
Then for Semifinalist, they calculate out a % allocation of SF’s based on number of high school seniors in each state, and 16,000 (which is the number of students they need for SF). Likely very close to the numbers from last year which can be found here: http://www.nationalmerit.org/annual_report.pdf. Then they take that excel spreadsheet and break it up by state and sort until they get as close as they can to the number of students they are supposed to allocate to that state - and presto. They have state cutoffs.
All the college board math that we are trying to extrapolate info for is not important for NMF. Only for parents who are trying to figure out if their kid is going to make $$$ in this contest.
He’s not just using one set to refute the other though. He is also using what he considers to be a significant amount of anecdotal data that suggest students are scoring in the top percentile at a rate that can not possibly be correct. Others have reported the same thing… too many students in the top percentile.
Now his speculation on what is going on is just that, speculation, and who knows if he will turn out to be correct on the details. But he has good reason to suspect that it’s the percentiles that are off - high schools are reporting dramatic spikes in the number of students scoring in the supposed top 1%.
He’s right, the percentiles do not pass the sniff test. While that doesn’t prove anything, it’s more than just using the concordance tables to refute the percentiles.
So CB has the real % data and they have % data based on a study…obviously it would be simple to see if the study results pass the smell test, but they ignore that and publish wildly erroneous information. Not sure I buy that.
bravo @suzyQ7 your message should be made a sticky and everyone must be forced to read before posting a message!
@YZamyatin I think what’s going on in Applerouth’s reasoning is that his anecdotal school data is compared against SI %tiles drawn from the entire pool of 3.5M PSAT test takers and not just Juniors. Obviously, juniors are expected to do better than 8th-10th graders who take the test, so no wonder they show up in top 1% of students. CB just needs to confirm what the SI %iles in pg 11 come from.
@YZamyatin. I understand Applerouth’s reasoning when he says that percentiles are inflated in their presentation. The National percentiles seem to be higher than the User percentiles. We can all speculate why CB decided to present the higher number more prominently on the report. But also we are all smart enough to realize that the User percentile is more accurate.
However, Applerouth goes on to make a big deal about more percentile discrepancies–those between User and SI percentiles. While he is right to explain the difference between the two, CB has done nothing wrong or suspicious in this case. Apparently the NM company wants to give out NMSF based on three evaluations–1)math, 2)reading and 3)writing/language. CB has decided to grade their college entrance exam on two evaluations–1)math and 2)reading/writing/language. Thus, one’s PSAT score and one’s NM score Are, in fact, Different. And, therefore, their percentiles are also different.
Once schools realize that the relevant PSAT score percentile is the User percentile, do you think they are still confused and reporting the number of high scorers to include both those students who scored 99% on the PSAT and those students who scored 99% on the NM scale? Couldn’t this be the reason for a perceived increase of high scorers?
So the National % numbers, User % numbers, SI % Table and concordance tables are all based off of a research study. And the tables disagree with each other …
Maybe next year, the 4M students can just skip taking the test and CB can do a research study to tell each student what score they would have gotten had they actually taken the test.
@Speedy2019 That’s a great idea - I think so many students & families are so tired of all the testing and problem that a virtual test or phantom score based on some “research sample” might be a better use of everyone’s time - especially given the way the scores / percentiles are developed, the various delays & also the way NM awards the SFs (different standards for states) which have really caused a drop in confidence in the CB. Also, I hear some feel that the CB’s suggestions on students PSAT reports about taking certain AP courses is money driven and not really student centered. For fun I just looked at my son’s - he is supposedly ready to take everything and when I put in his planned major (Elec. Eng), the list of recommended AP’s includes Music Theory!
@PAMom21, Interesting. I was hoping to be evaluated for national merit in the state of Nebraska since their cutoffs seem to be lower. Anyways, I guess I’ll just have to see how it plays out. Do you know if the person you were talking about was actually a Colombian resident? Or did they maintain residency somewhere in the U.S.?
@Willis1313 not sure if you got a complete answer to your question but I just found the “Official Student Guide to the PSAT/NMSQT” for the class of 2017 and here is what it says:
“A participant can be considered for Semifinalist standing in only one state or selection unit, based on the high school in which the student is regularly enrolled when taking the PSAT/NMSQT.”
@Mamelot, Awesome! I had looked through some different sites but I was having trouble finding exactly how they determine the selection unit. Thank you so much.
Yes - this sounds right. What I am not sure is if the top 50k scorers represents a “hard line” or fluctuates. I think it moves a little. Even so, whoever scores within this range is by definition “commended” but also may qualify as NMSF depending on their state. What I do not know is why the number is generally around 50k. This does not represent the top 1% of scorers but the top 3.3%. I had thought that this 50k cap exists b/c that’s what it took generally speaking for low scoring states to fill their allocations. In other words, NMSC had to go down to the kids scoring at 98/97% nationally but whose state score is their top .5% to get them to qualify. But maybe I’m wrong. The reason I ask is that if this the 50K cap stays constant and 1.5 mill took the test, then the commended line has to be around 97% and not the 99% line Testmasters is saying it is. But I’m not a stats guy and maybe someone can explain how 50k kids can = 1% of a test taking population of 1.5 mill.
It seems most on here believe CB squeezed 3% of the test takers into the 99% in order to get a one time bump in folks feeling they need to take the SAT instead of the ACT. I can’t believe they’d burn their reputation for so little return.
If 210 SI in fact is the commended cutoff as some are speculating and the prelim concordance tables are basically sound, will CB
(1) Issue “updated” score reports indicating the real user percentages to all the test takers? (probably not)
(2) Issue an updated guide to understanding the SI and national/user percentages that will show the re-calibrated percentiles against the real test population? (possibly)
@suzyQ7 , Power to You (post #4329)