@gmtplus7
You’re correct; score percentiles are meaningless, but this has been stated numerous times. It’s dependent on the SI, but our issue is we don’t have a good way of correlating 2014’s SI and 2015’s indices.
I also agree with @GMTplus7 and @EarlVanDorn analyses of the process. I also agree that percentiles are a red herring, @EarlVanDorn. I don’t think there is enough reliable information to accurately assess the percentiles down to the tenth of a percent before the state reports are released (and maybe not even then). And since the number of test takers per state varies year-to-year, drawing a correlation based on percentiles without the state reports is not helpful in my opinion.
@ILfather The number of test takers doesn’t matter, the proportional allocation of the 16,000 semifinalists is based on the number GRADUATING SENIORS in a state, not test takers. The only impact more test takers have per state is if those additional test takers are high scorers. I can’t see that being likely (we know there are additional juniors who took the test this year because it fell on a weekday, but what are the chances that a large proportion of them are high scorers?)
@ILfather , @suzyQ7 has it right, the percentage is based on the number of high school graduates, not the number taking the test or the number attending school. But even that isn’t precise, because one index number might result in 1.2 percent of the people in the state getting SF while a point higher might result in only .9 percent. So they have to do a lot of playing around with the numbers, and in the smaller states they can jump up and down quite a bit.
My state has a graduation rate of 75 percent. We just did away with subject-area tests as a requirement for graduation, so no doubt that percentage will increase as more students graduate. I say we just give everyone a diploma and increase our number of NMSF by a third! (or is it a quarter?).
NJ, 750 M, 680 V, 1430 T, 99%, 211 SI
Do you guys think I have a shot at being a semi finalist?
- 620 math, 720 writing/reading. Selection index 206 - do you think this will be enough to make commended in California?
FYI, My calculations in post#114 plotted in a table alongside the College Board’s PRELIM concordance table:
http://talk.collegeconfidential.com/discussion/comment/19169349/#Comment_19169349
WV
Total 1490
Reading/Writing - 730
Math - 760
R - 35
W - 38
M - 38
Index - 222
Received TASP email and Stanford letters
Illinois
Total 1460
Reading/writing-720
Math -740
R-37
W-35
M-37
SI -218
No emails or letters for TASP or otherwise
@EarlVanDorn, @suzyQ7 – Mea culpa. You are absolutely correct – # of semi-finalists is based on graduating seniors, not test takers. My apologies for mis-stating that.
I appreciate the conversation, and really like what @GMTplus7 has put together. Here it is again in a slightly different format:
Now I am officially OLD (“senior”) growing a few more white hairs battling CB incompetence.
@ILfather Thanks for the chart. Can you help me interpret with a real example? Does the chart say that if you got a 1480 and 221 SI in 2015, it equates to a 227 in 2014?
Massachusetts
Total - 1470 (99th)
RW - 740 (99th)
M - 730 (99th)
R - 37
W - 37
M - 36.5
SI - 221
No emails.
Ohio
Total - 1480 (99th)
R+W - 720 (99th)
Math - 760 (99th)
R - 34 (95 user %tile)
W - 38 (99 user %tile)
M - 38.0 (99 user %tile)
SI - 220
I did get a TASP email and tons of emails from colleges.
Good question, @suzyQ7. My interpretation matches yours, but @GMTplus7 can offer a better perspective.
@suzyQ7 I think it corresponds to 227 to 231 range under the old 240 format. No exact number.
@websensation, I agree your observation is more accurate.
Actually, on second thought, I suspect a 1480 (221 S.I.) implies a range of 225-227 in the older formatted PSAT.
<pre><code> ratio of 2015 S.I. /
</code></pre>
2014 | 2015 || 2015 PSAT | 2015 PSAT S.I. | 2014 PSAT score
227 | 1480 || 1480 | 220 - 224 | 0.969 - 0.987
226 | 1480 || | |
225 | 1480 || | |
