"Semifinalists must advance to Finalist standing in the competition by meeting high academic standards and all other requirements explained in the information provided to each Semifinalist. Click here to learn about requirements for becoming a Finalist in the 2016 National Merit Scholarship Program. (Adobe Acrobat Reader is required.) back to top
Finalists
In February, some 15,000 Semifinalists are notified by mail at their home addresses that they have advanced to Finalist standing. High school principals are notified and provided with a certificate to present to each Finalist. back to top
Winner Selection
All winners of Merit ScholarshipÂź awards (Merit ScholarÂź designees) are chosen from the Finalist group based on their abilities, skills, and accomplishmentsâwithout regard to gender, race, ethnic origin, or religious preference. A variety of information is available for NMSC selectors to evaluate: the Finalistâs academic record, information about the schoolâs curricula and grading system, two sets of test scores, the high school officialâs written recommendation, information about the studentâs activities and leadership, and the Finalistâs own essay."
@Tgirlfriend Iâlll try to get the excel out tomorrow⊠had a busy weekend. @srk2017 Post 2189, you heard a few of 218âs? How many? Iâd like to put this in my excel. CA right?
From my understanding here are the Means for the research study group and actual test takers:
Research Study Reading Mean 489
Research Study Math Mean 498
Research Study Total Mean 987 with a standard deviation of 174
I believe that comes out to a Mean SI of 147.6
Actual Testers Reading Mean 507
Actual Testers Math Mean 502
Actual Testers Mean 1009
I believe that comes out to a Mean SI of 151.6
So the average score is 22 points higher for the actual testers, and the SI is 4 points higher for the actual testers over the research study group.
I think someone calculated that the 99.5% for the study group was 214 matching the 99+ designation in the Guide. Is that right?
What about the actual testers, I donât think we know the standard deviation for that group, right? Can someone estimate the 99.5% for the actual testers?
Can we do 214 + 4 and say 218 is 99.5% for the actual testers as an estimate?
@rb681000 In response to your #2218 this is what I found outâŠhope this helps.
My daughter has a 212 SI in TX. She took the SAT in Oct and Nov 2015. We are happy with her SAT scores so we are done! However, her 212 potentially could put her at NM level but we wonât know that until Sep 2015. I was concerned if we found out in Sep then we already missed our chance of sitting for the March, May 2016 exams. We donât want her to take one more SAT unless she absolutely has to.
So, I called NMSC and explained my dilemma. The gentleman in administration explained to me the qualifying SAT score process.
Their confirming qualifying score process is a 2 year process. In her case, since she took the PSAT in Oct 2015, they would take any SAT that she took from 10/2014 through 12/2016âŠidk why Dec 2016 and not Oct 2016âŠclose enough, so I didnât question him. Oh, they would take the highest score from those tests.
The score has to be good enough. In the past 3 years, the score had to be at least 1960 and he doesnât see a change for this year (for the old SAT). However, they donât have the qualifying score for the new SAT yet.
He went on to caution me that if my daughter made NMSF, be sure to have CB send score over to NMSCâŠitâs not an automatic process between CB and NMSC.
So, for schools that principal/GCâs donât pay that much attention to NMSFâs, donât rely on the schoolâŠgo straight to NMSCâs website info and call about qualifying score. NMSCâs website has good info about advancing from semi to final but not all the specifics are there⊠at least in my case.
@Speedy2019 - GeorgiaâŠ218âŠ210-220âŠ216âŠ210âŠ220
I agree with the higher cutoff in your table. For example, I believe GAâs cutoff will be 220 or even 221. I am from GA and did my calculations a while ago based on the top 3 schools in Cobb County (Walton, Wheeler and Lassiter). I am no statistician. Just common sense. Walton isnât the best high school in GA. There are at least 5 high schools just as good or better than Walton. Since Waltonâs top 100 studentsâ mean score is so high, I would venture to bet at least a 220 for the state cutoff.
I finally read the Compass Prep article, âProblems With the New PSAT Part 1â that so many of you have been recommending. This seems to be the explanation we were looking for. The article explains how using the alternate definition of percentile, that is, the percent of students scoring at or below your score (instead of just the percent scoring below your score) affects the number of students contained in each of the percentiles:
âAt the edges of the scale, the absolute change is smaller, but the proportional impact is higher. For example, while 99th percentile is only 2 percentile higher than 97th percentile, the cumulative reporting changes from 2014 mean a doubling or tripling of students receiving the higher figure. With 3% of students boasting 99th percentile scores this year, there are important implications for how students, parents, and counselors forecast National Merit scores, for example. Counselors typically see hundreds of reports, so they have been observing this proliferation of high scores without necessarily knowing why.â
This article is explaining that if the old definition of percentile had been used, we would probably have percentile tables that made sense to us. The number of students with scores designate as being at the 99th percentile would probably have been similar to that of previous years. But since the definition being used adds up the number of students making the same score as well as the number of students who scored below that score, scores are being designated as 99th percentile that would not have been 99th percentile previously. They would have been at a lower percentile.
I am not at all sure how to estimate the change to the upper percentiles as a result of the actual testersâ means being higher than the research study but it seems plausible there will be a shift in the score distribution; impact at the 99-99+%ile level is less clear but certainly possible too, right? Maybe Doyle B & @thshadow can help - @Plotinus are you still out there? thanks!
@CA1543 Sorry, I have been very busy writing new lessons for the redesigned SAT and havenât had time to follow the NMSF discussion that closely. I did take a quick look at the CB validity study report. I noticed that the first-year college students in the study were offered a monetary gift card as a reward for exceeding their highest scores on the old SAT. This was supposed to motivate them in the same way that students taking the real test are motivated.
I find this ludicrous. Does CB really think that a small GIFT CARD is going to motivate students the same way that they are motivated by college admissions verdicts and financial aid awards?
Another thing that was strange was that there were twice as many girls as boys in the research sample. Now what is that supposed to mean? Shouldnât the research group have the same proportion of girls to boys as does the actual test population? Or are we loading dice here?
The sniff test told me back in September there would be more high scorers than there had been before. Now that I have studied the October PSATâs in detail, I can confirm that the test is way, way too easy. Some questions are so easy itâs embarrassing. The questions that are âhardâ are hard in the wrong ways (tricky language, trap answer choices, not conceptually challenging). I agree it is embarrassing that most US teenagers cannot do division of decimals without a calculator, but how much does that really matter? That is the difference between calculator-habituated and not calculator habituated. The ability to do 300/0.06 without a calculator should be the difference between A and C in fourth grade arithmetic. That should not be the difference between NMSF and not NMSF, or between admission to or rejection by Harvard.
Yes, I think the test is really dumbed-down. ETS, I miss you.
Itâs too bad, because the idea of aligning the test better with the most important skills needed for college is a good one. However, I suspect that somehow, many students unable to compute 300/.06 without a calculator are thriving at Harvard.
That said, I still donât think there is enough data to draw any definite conclusions about specific NMSF cutoffs in specific states. Those are going to depend upon what is happening locally in different areas, not what is happening overall.
@SLparent â thanks again so much for continuing to work on your chart - nothing like real data. I will look through some other threads and see what we can find about other PSAT scores.
@BunnyBlue#2249 I am not sure if what this is article is saying about the definition change is entirely correct. Posters here have pointed out that the definition change at the 99% level is negligible, a point shift. I take this to mean that if the 99% range should approximate 16k testers out of a population of 1.5 testers, this definition change would impact or affect about 1.6k testers in terms of their actual ranking/placement, not 2% of the test population (+ - 34k testers). But maybe I misunderstood.
@CA1543, âI am not at all sure how to estimate the change to the upper percentiles as a result of the actual testersâ means being higher than the research study but it seems plausible there will be a shift in the score distribution;â
Yep, the actual testers did better than the research group. I donât know any way to estimate how much that affects the 99+ group.
I would say the 214+4 = 218 for 99+ would be the outside extreme.
So probably between 214 and 218 for 99+. I know⊠doesnât really narrow it down. The 218 though jives with some of @DoyleB thoughts on a potential local peak in that area.
I think 2015 was more reasonable, but I lived through 2014, and Iâm pretty sure it was the 28th of February, . I remember thinking they barely kept that promise. Maaaybe the 27th.
Yup, editing to verify the 28th. Hereâs that discussion. But I remember being surprised the following year that they were out so âearlyâ (itâs all relative).
I enjoy following what you all have to say. I have learned so much. You all are forcing to sweep some major cobwebs out of my brain. Anyway, I find it so interesting what @Plotinus posted that their were twice as many girls as boys in the research study. That really makes me question how careful they were trying to be when setting up their research study. I recall, and you all would probably better know the research on this than I, that there is a significant difference between girls and boys and how they preform on standardized tests. Obviously, girls generally better at reading and boys at math. But I remember something to the effect that girls perform better as a whole but boys are disproportionally represented at the extreme high end, like the top 1% that we are all trying to figure out. It would seem very irresponsible for CBâs research sample to not accurately reflect the gender breakdown of the actual test takers. @Plotinus I also agree that the test seemed far too easy. It seems that this new redesign would be easier than ever to prepare for because there are so few conceptually difficult questions. My seventh grader could have done as well. And I am not saying that to brag about how smart my seventh grader is but these concepts would have been introduced at least by middle school at least. That is probably why their is not as much difference between the grade levels as previously. I worry that the students that will do the best on this type of test over time will primarily be the most meticulous and careful. While being careful is important, I donât know if it the best predictor of who will really thrive in college and how we should select NMSF.
. I remember thinking they barely kept that promise. Maaaybe the 27th.