National Merit Cutoff Predictions Class of 2017

SI % TABLE is from research study: https://twitter.com/sofferga/status/688863927837528064

This makes sense now. They don’t use real SI percentiles because we would be able to figure out the NMS commended cutoffs with that information and could probably infer the rest. They have never used the real SI percentiles.

Actually I believe they have published the real SI percentiles in the past - they just publish them one year after the fact, when the cutoffs. etc. are already known. That’s why when you want to know where in rank order your state’s cutoff was, you need to look at the “Understanding the Scores” document from one year later.

The bullet on the SI table of previous “Understand” documents reads:

“Percentiles are based on the Selection Index earned by college-bound juniors who took the PSAT/NMSQT in the previous year.”

I assume next year they’ll publish the real table from this year. In the first year of a newly designed test, they don’t have real data to publish from the previous year, so they use “research study data”.

They have the actual data and could easily conduct a reasonableness check on the study results prior to publishing. I just find it hard to believe that they would publish a wildly inaccurate table.

The problem now is that we don’t know whether the “research study” for the SI table is based on users or 11th graders in general. Big differences between the two.

so in other words the percentile and the concordance tables are useless. Thanks College Board


@DoyleB since current document does not say anything like that, my guess is that the percentiles for SI on page 11 is based on the entire pool of 3.5M or so test takers, and not just 11th graders. They also can’t use previous year’s data because they would have to use concordance themselves! They could use a research sample (whatever that means). In prior years (2014) if they based percentiles from previous year’s test results, this year it’s harder to do so because 2014 test was on a scale to 240 and 2015 is on a 228 scale. None of this is pertinent even to NMSC, as I think what matters is what @suzyQ7 has to say in her message #4329 at http://talk.collegeconfidential.com/sat-act-tests-test-preparation/1816872-psat-discussion-thread-2015-p289.html

At the top of each column on this year’s SI table, it says:

SELECTION INDEX
PERCENTILE
(GRADE 11 ONLY) "

I think the SI % Table is for 11th graders only. But still could be “National” or “User”. USER makes sense to use, but you never know.

Ugh I asked my GC how many students scored higher than me on the PSAT and she said she can’t tell me other students’ scores. It seems weird that things like Naviance are OK at other schools because that would be a broader violation of privacy in my opinion. Maybe in OR we are a little libertarian- idk haha.

I stated before that if the data was “national”, then they’ve gone from 1.5 million to 3 million students. If they assume that kids who didn’t take the test previously were likely to have scored lower (at least below the top), then they add those kids to the lower parts of the distribution. That effective halves the distance from the 100th percentile for kids in the top of the chart.

Another way of saying this is that, if the chart is “national”, to convert to “user” you would make the bottom half of the 99+'s into 99s, and the 99’s become 98s. When you do that the percentile table starts to make more sense, the concordance tables make more sense, and the number of kids that might make NMSF makes more sense.

@Speedy2019 - you quoted the question about SI percentiles - but not the answer. I assume Collegeboard actually answered you, but your twitter quote failed to include it?

@DoyleB I understand your logic, but you just said 11th grade only. So even with national, you are talking about 3 millions JUNIORS??? Just clarifying.

@PAMom21
I now have access to the real test questions and have gone through the non-calc math questions.

I am really shocked how easy they are. They seem easier to me than the practice test questions. I admit someone who can’t do arithmetic of fractions and decimals without a calculator will be in trouble, but we are talking about very easy arithmetic of fractions and decimals.

Here are some of the hardest operations that you have to do:

  1. (.06)(30000)
  2. 120/.06
  3. [(2/3)(b)+b]/b
  4. solve the system 56x+35y=2604, x+y=60. Here you are going to get big numbers unless you reduce the first equation to 8x+5y=372 or use the elimination method. But even if you don't reduce or eliminate, it is still doable. If you do reduce, it is easy.

The calculations are middle school stuff at best. I don’t see how any student with an A in math had trouble with the numbers.

@thshadow

Speedy2019’s link showed the following response:

We are in Texas also and trying to figure out if a 221 will get in the SF or not. The next closest score in my sons school was a 202 and she is right behind him in class rankings. @SLparent

@Tgirlfriend Although no one can say for certain, 221 in Texas is likely to make SF.

@payn4ward 
that is what we are hoping for however I keep seeing all the posts about people thinking there is going to be a lot more perfect scores and it makes me second guess everything. September is a long ways away from now!!! There needs to be more time to prepare and to plan once the SF list comes out. Just my thought.

@Tgirlfriend Sorry, there are just too much uncertainties. There is nothing we can do till September to find out. Some people have not even received their PSAT scores!

Commended cutoff is expected in April. It will NOT help at all with Cutoff estimation but we will know how much off CB research percentiles are (around 97%) !