We have changed the way we log in on College Confidential. Read more here.
National Merit Cutoff Predictions Class of 2017
This discussion has been closed.
Recent Activity
-
4.4K 106
-
177.5K 1.1K
-
220 2
-
831 33
-
1.3K 23
-
3.6K 35
-
109 4
-
13.5K 147
-
81 0
-
446.5K 1.1K
Replies to: National Merit Cutoff Predictions Class of 2017
"2. When I convert my scores using the concordance tables, I get a score nowhere near the top 1% (206 out of 240). However, I scored a 1420 (211si) which puts me within the top 1% by about 30 points composite (1390s have reported being in the 99th). Others have already detailed problems with these tables, so it's safe to say we shouldn't put much trust in the charts."
You didn't say what your actual scores are so it's hard to follow this.
"3. If I convert by using my percentiles (99% in math and verbal) to the percentiles on the old tests (meaning that I took my 99th percentile in both math and verbal and converted to what a 99th percentile would have been on the old test it converts to a 222 out of 240 (and that is using the lowest possible scores to qualify as the top 1%). So, is it possible that converting your scores this method provides a more accurate idea on how this converts to the old tests? "
One thing that complicates using percentiles is that we know that a lot more students took the PSAT this year--I think 1.5 million more. Those were students who presumably didn't want to take the PSAT and were forced to do so by their schools. I can tell you that in our school they make all kids take the PSAT, even kids who aren't planning to attend any college. So, how well do you think this new group of test-takers did? I'm sure there are some great students in that group but I would guess that this group populated the low end of the scores while making it easier by their sheer numbers for good students to get into the 99th percentile. More test takers = a greater number of students in the top 1%. So while I do think you can get some idea by looking at the percentiles, for high scoring students, I think your percentile on the new test will be higher than it would have been on the old test.
Personally, I am inclined to believe the concordance because the college board has all the data--we don't--and I assume they have some statisticians who have been over this very carefully. But maybe they just made something up and slapped it up on their site. I only looked at a few of the top scores on the concordance because I was mostly interested in how the high cutoff states may have changed.
Now you need to back it up with the SAT and keep your grades up ;-)
Really, there is no excuse for the college board--after a full extra month creating the super duper detailed score reports, they left out the information in that everyone wants to know (eg. a complete chart of scores and the percentiles they correspond to, and the state cut-offs for NMSF). I don't think anyone cares about their fancy but uninformative score reports.
Will a 1470, 221 SI make semifinalist in CALIFORNIA (CA)?
I scored in the 99th (just barley) on the verbal, 690. The lowest score still in 99th percentile on the 2014 PSAT for CR and writing was a 73. So let's assume that means that I got a 73 on both sections. I could have scored lower writing, for example, but the higher reading would balance that. Then on math I scored a 730. The lowest possible math score on the 2014 PSAT that would have still kept me in the 99th percentile was a 76. Adding 73+73+76=222.
Last year (2014) about 1.5 million juniors took the exam. This year we know that about 1.5 million juniors are eligible for NM. Although, not all juniors from 2014 are eligible, a large percentage was. So, I think that although way more kids took it last year, people are blowing it out of proportion how many more juniors will be in the running. More? For sure. However, not a million more kids as people seem to be reacting.
The concordance tables though simply don't match up percentile wise. I scored my 1420 converts to a 206. 206 was the lowest possible score to be in the 98th percentile in 2014. However 1420 is actually safely within the top 1%.
First of all, your scores would not be in the 99 percentile if that 1.5 million students had not taken the exam this year. Would be around mid-98 percentile. Not sure how the college board handled this influx of students but if they are trying to tell you what you personally would have scored, it may be that the percentiles don't like up the way you are assuming they will.
Secondly, you cannot convert the math plus verbal and compare that to a selection index. They weight the math and verbal differently. On the selection index, reading, writing, and math are each 1/3. On the math plus verbal, math is 1/2 and reading and writing are each 1/4. You are stronger in math and weaker in verbal, so of course your selection index is a lower percentile than your math plus verbal is,
What happens if you look up your verbal subscores instead of the composite? Are the subscores also in the 99 percentile?
I understand what Jay is trying to determine. I had 3 kids take the test. Based on their results, the lowest 99th percentile score for Reading is 35; the lowest for writing is 34 (so 690); and I believe the lowest for math is 37 (because I had two kids score 37.5 and they were 99th percentile for "users" and one kid score 36.5 (that child was 99th percentile nationally but not 98th percentile for "users").
I reviewed old scores from 2010, 2012, and 2013. If you scored in the 99th percentile for each INDIVIDUAL subject test (and added those scores together), you would qualify for national merit in almost every state, except for the VERY highest.
So, I think what Jay is trying to do is say since I was in the 99th percentile in EACH INDIVIDUAL subject, my total score should correspond with at LEAST the lowest score for the 99th percentile in each individual subject on prior tests.
However, my son's scores are at the 99th percentile for W and R and just half a point off for math, and his concordance score is 213, I believe, which is no where near the >220 that I calculate using the old scoring for each subject at 99th percentile. The concordance is on target for my two girls' scores. So, maybe it goes off track or maybe I just don't understand it.
Also, on another point the page with the selection index still says that about 1.5 million eligible students took the test.
I know everyone thought the test was easy, but I am confused in that being in the 99th percentile in every subject test is the 99th percentile, right? No matter how easy people thought the test was.