|
Today I found out that my DCPS junior got three wrong out of about 120 questions on the PSAT, apparently missed being a NMSF by one question. Still in the 99 percentile (top 1%) in the country, but does not make NMSF because of DC's cutoff, highest in the country.
Why is the cutoff so high? |
A better question might be why is the test so easy. Seems to me there ought to be a bigger spread between top scorers and the rest; otherwise starts seeming like luck (rather than pure smarts) is too much of a factor. |
The cutoff is high because only .5-1% of students who take the exam are supposed to become NMSF. There are many kids in our highly educated area that score well, and miss fewer than 3 questions. |
Because DC is not a state. If you take the test somewhere that is not part of a state you are automatically assigned the cutoff of the highest state. |
This is a problem in general with standardized testing, the tests aren't hard enough to give much information at the high end. My son interviewed for MIT the fall of his senior year, after taking BC calculus as a junior. The interviewer asked if he had taken the BC AP, and then said, "I assume you got a 5?" That year 120,000 kids got a 5. |
|
FYI, other states have an acceptance rate well over 0.5% to 1.0%. Check out page 2 here:
https://www.nationalmerit.org/s/1758/images/gid2/editor_documents/annual_report.pdf?gid=2&pgid=61. |
And when that happens they almost always raise that state's qualifying score the subsequent year. |
Yes, and this is one reason that International Baccalaureate Diploma exams have steadily become more popular in this country in the last several decades. IBD exams come in two forms: Higher Level (one or two years past AP study) and Standard Level (roughly comparable to AP, but without multiple choice questions). IBD exams are particularly useful in separating the sheep from the goats on foreign language proficiency and writing ability. Scoring 7/7 on a HL IBD language exam is next to impossible if you aren't a native speaker, or close. By contrast, more than 80% of the takers of some AP language exams, e.g. Chinese, score 5s. |
The 2019 99th percentile ranges from 1370-1520. |
and accounts for over 17K students. |
Only if you look at the number taking the PSAT. In DC, 95% of juniors take the PSAT, in Alaska it is 21%. However, NMSF is apportioned based upon the number of students in the state, NOT the number taking the PSAT. So, Alaska had 37 semi-finalists out of 2000 test-takers and 10,000 juniors => 0.4% like all other states DC gets 35-55 per year out of 5500 students. => ~0.8% Look at the state SAT suite of assessment reports. |
>4% of the 1.8M PSAT takers got > 1400, or 70,000. The Nationally representative percentiles are not real (and based upon 3 prior years). See the CompassPrep blog for more info on how "Everything we think percentiles tell us about National Merit is wrong." |
Because the PSAT is geared toward the average student and is poorly suited to distinguish between the 0.2% student (1500) that makes NMSF in DC and the 0.3% student (1490) that does not. |
And also explains why most kids just don't bother, and focus their time and energy on things that will matter on their college application. |
|
I also heard that the cut-off in DC this year was 2 missed questions. So the DC semi-finalists missed either 0, 1 or in a few cases 2 questions (questions are not all weighted the same so some who missed 2 did not make it as a semi-finalist). In contrast in many states kids could miss 10+.
So no way is that a reflection on DCPS. To miss 0 or 1 questions a kid has to both know the material and be a phenomenal test taker. Zone out for a minute? Your score is blown. I wonder in part if the environment the kids took the test effected their scores. Wilson gives the test to everyone. You can imagine those PSAT classrooms could be chaotic. If the standard is perfection or 1 question from it, a difference in environment (quiet testing room vs. not) could easily affect who misses one question vs. who misses two. |