About the “different populations”: This is exactly what I meant when I wrote about the “hot potatoes” involved in the curve. The hot potatoes are the populations. For decades, the SAT has been accused of being a test that is biased against:
lower income groups
racial minorities (blacks , Hispanics, and Native Americans)
females.
It is going to look bad if the scores for the new test are worse for these groups than they were for the old test, or if these groups performed differently on the two administrations.
About different tests: There have long been two completely different PSAT tests each year, one administered on Wednesday and one administered on Saturday. No two tests are identical in difficulty level. CB uses different curves for different tests to compensate for differences in difficulty level. This is what “equating” is.
One issue could be adjusting the curves (or enhanced curves) to make sure that the above-mentioned populations performed the same on both administrations, or the same as they performed on the old-format PSAT.
About “throwing out” questions to adjust the curve: I think this is really unlikely. Student will get back their original test booklets in January. Everyone will know what the original questions were. The only valid reason for throwing out a question is that it has an error in it. CB would have to show that the question has an error if it wanted to throw the question out.
The only acceptable way to adjust a curve is by changing the correlation between a given raw score and a scaled score. The difference between -10 and -11, for example, could be 0 scaled points, 10 scaled point, 20 scaled points, 30 scaled points, etc., depending upon the curve, and so on for all other raw scores.
Thanks for this explanation, @Plotinus! This all makes sense. I would think they need to provide state populations to NM so that the latter could establish cut-offs, etc. and I know that there are national merit programs for various ethnic groups. So my question would be whether the “Index” provided to NM is based on an overall percentile or somehow “normed” for gender/ethnic group etc?
@Studious99 my thoughts exactly!! NM will need to update it’s information about how 34,000 are Commended and 16,000 become semi-finalists. With 4.5 million test takers entering the NM program automatically (PSAT is first cut) top 1% or so are going to be around 45,000 kids! Depending on the overall population that should open up National Merit designation to some kids who wouldn’t have been included had they taken the old test.
@Mamelot there cannot be norming by ethnic group, as that information is not required at test time. There is s separate National Hispanic program that is scaled based on that groups scores, but that can’t possibly be the case for the general NMS program.
Just because more students are taking the test doesn’t not mean that there would be a huge increase in tippy-top scores. Students have always more or less self-selected the PSAT (especially in public schools) only when they think there is a chance of a good score. Most others don’t bother taking it. If schools are funneling more of their population to take the test, that doesn’t mean they are more likely to get great scores than the kids that are motivated to take it.
Agree, partially, but if 3 million ,“extra” kids take it, there must be quite a few high scorers in that pool that were forced to take it, that may have self selected themselves out in the past, or didn’t take it in the past because of sports commitments. 3 million extra is a HUGE number, huge difference from past tests. There is no way that all 3 million are in the bottom or middle of the pool.
Of course this is speculation but some of those 3 mil are going to be in the top and some of them are going to do great and wonder what the big deal was about this test!!! I think that was probably CB’s purpose.
I wouldn’t exactly say the test has been dumbed down because it’s all relative. Whether the test is easy or hard you will have tippy-top over achievers setting the bar pretty damn high. However, CB is hoping to find more of those overachievers in the 3 mill. pool and the numerous comments about how easy the test was is a testament to that.
It really troubles me that I cannot find the questions and answers for the 2014 test that used to be available on my son’s College Board page. He was planning to look at them as he practices for the January SAT score. Have they been removed? Or just moved? (To where?)
I think they have been removed when CollegeBoard updates the quickstart website. If your son wants to practice for the January SAT, I recommend that he finds some old SAT tests and practice by doing those in real time.
You guys have made my head hurt with your thorough analysis! Ok, so assuming scores actually come out in January, when do you feel we’ll have a good idea of the NMSF cutoff? Do you think we will be able to compare new selection index to previous years? I ask because we, too, don’t want to send CB any more money than necessary and may just stop our standardized testing at the October ACT.
Just wondering, is the National Merit Finalist recognition actually based on the top 1% in the state? If so, more students taking the test should benefit the high achieving students as the over-all number of finalists increases: many kids I know, particularly from public schools where they were “forced” to take it really just blew it off and did not even try.
Their student guide says there will be 50,000 students in the commended (34,000) plus semi-finalist (16,000) pool–not a specific percentage of exam-takers.
“About 3.3 million students are expected to graduate from high school in 2015–16, including 3.0 million students from public high schools and about 0.3 million students from private high schools.” (source: NCES “Fast Facts”)
So, assuming that the class of 2017 isn’t all that different in numbers and make-up, that 4.5 million student number of PSAT test takers must include kids who are not going to be in the NM program - either foreign students or non-college bound kids.
But I find it very hard to believe that nearly 100% of juniors took the PSAT this year. So some of that 4.5 mill also much include PSAT 10 and PsAT 8/9. What’s left over after subtracting all those sub groups out? I’m predicting a larger number than the normal 1.5 million pool but surely not double (or triple!).
@workinghard1234 it might be similar to what goes on with the “required” ACT (which more and more states have adopted as part of their grad standards). In reality - at least in our state - no one is “required” to take the test. Rather, districts are required to administer them “free of charge” (i.e. taxpayers pay). But the student might be able to opt out. My D2 declined to do the ACT last year when it was newly incorporated into our state standard. The wording of the rule allowed for the SAT as well (it may have said ‘similar standardized test’ or some such). She was prepping for the SAT and that’s the test she wanted to use. She got a pass on that no problem. I’m wondering if students and families realize they can obtain the actual wording of the state rule and possibly opt out to take the test of their choosing.
Ok. I just went and checked my email got the Brown letter. HOWEVER it has my college Juniors name on it because he used the same email(mine) when he took the PSAT. So now I’m not sure if they released score information or not.
Because for the SAT my junior used a different email, however he did use the same one for the ACT. Anyway it all seems pointless because I think they are going to do something funky with the scores to get the numbers they want. Remember this is the same company that said it was fine to not score all sections of the test and still have a valid score.
Of that 4.5 million, how many are juniors? That’s the only relevant data for NM.
My kids have been taking the PSAT since they were freshman to practice it, and younger D15 took it this year as a sophomore. But her score doesn’t “matter” in terms of national merit.
Our public school pays for all sophomores to take the PSAT, although they don’t require it.
As to who’s getting the Stanford sticker-both kids have taken the SAT/ACT at age 12 (for TiP), and both have taken several AP’s. D15 did not get the Stanford sticker, and the only difference that I can see right now is that she did not take the Dec 5 SAT.
I can’t remember if the Stanford summer school thing came before or after Dec 5, either. It’s just such a flood of mail all the time.
Here’s another thought: maybe the huge jump in numbers of kids taking the PSAT this year and the delayed scores are a precursor to having some sort of merit scholarship award becoming available to kids in freshman and sophomore years. It would explain why so many more kids are taking it IF the big jump is from the non-junior years.
And if that’s true, they may be applying all sorts of magic to those younger kid numbers, and that’s what’s holding things up.