Will the National Merit Semi Finalists be the top 1% of test takers based on the Nationally Rep. Sample or the PSAT/NMSQT user percentile? My DS is in 99% of the Nationally Rep. Sample but 97% for the PSAT/NMSQT User percentile. He got a 1410 and 211 selection index. I can’t figure out how a 211 would be in the 99% but predictions are that 212 will be cut off for Commended.
The percentile that is of interest to you is the user percentile. It uses only people that actually took the test. The Nationally Rep sample is their guess of the percentile if everyone took the test. Since the top 50k test takers are normally the top 3% of the people that take the test, that is normally where the cutoff is for the Commended. Note however, because of the two different score methodologies, you can be in different percentiles for the PSAT and the NMQT. I am guessing that with a 1410 and 211, your DS did better on the math section than the reading/writing section. That is why he is on the top of the 97th for the Total PSAT but right at the cutoff for NM. It is possible for some with a lower PSAT to have a higher NM Selection Score (e.g. 1400 212).
The reason for the difference between the National and User samples is that only 1.6mm of the possible 3.6mm juniors actually take the test. The distribution of the people that don’t take the test is lower than the test taking population.