Did I miss a memo? I have not seen any reference to an “equating section” on the new SAT. Don’t they need one of those? What have you all heard?
It does not look like there will be an “equating”, “variable”, or “experimental” section in the SAT. I have been keeping an eye on Collegeboard’s communication, but don’t see anything that explicitly say so.
I do know that the new PSAT next month will not have an experimental section: https://collegereadiness.collegeboard.org/pdf/supervisors-manual-psat-nmsqt.pdf
However, the PSAT 8/9 does have an experimental section: https://collegereadiness.collegeboard.org/pdf/supervisors-manual-psat-8-9.pdf
Well that raises some interesting questions. How will they determine curves? How will they know if a weaker set of tests is a result of weaker students or a harder test? Also, if you look at the charts for the 4 released “practice” tests, the charts vary – as if they were adjusted based on information obtained from equating sections. Are these just made up?
Although actually, now that I think of it, they do have data on these questions - they have been sticking them on experimental sections for the past year. Maybe that’s what they plan on doing going forward: maintaining the pedigree so that there are always enough questions on a new test that link back to this initial data set. But that won’t help at all in assessing the level of difficulty of new questions. That being the case, I would expect that the new math test will not be as clearly ordered in terms of level of difficulty.
Lots to think about…
I am not sure how they will test and approve new test questions. The whole idea of the experimental section was to weed about questions that may be ambiguous or not well written. This is standard practice for the GRE and GMAT. How does ACT deal with introducing new questions? There is no experimental section in ACT. I don’t know enough about psychometrics to comment on these changes.
According to this https://collegereadiness.collegeboard.org/pdf/test-specifications-redesigned-sat.pdf
CB has been testing redesigned SAT’s with sample student groups from schools. Maybe that will continue after the test rolls out.
Thanks, @Plotinus
I’d be interested to hear more about where they are finding “motivated” high school students to take their sample questions. I teach plenty of very motivated students, but they are not so motivated when the tests don’t count. (And who can blame them?)
It’s also interesting to read about how they are reaching out to secondary school teachers to include them in the process. I don’t know if it is still the case, but the ACT used to do that – hire freelance teachers to develop test questions.
I agree. The whole thing looks like work-in-progress. Did I miss something or is the top score on the Official practice PSAT less than 800 per section? So what does that mean, the hardest questions were left out? The 4 practice SAT’s also look to me as though the hardest questions are missing. I couldn’t find a single question that looks like a 5. As the tests stand, it seems to me that a zillion people at the top would write perfect papers. On the other hand, the people in the middle might be in even worse trouble than they were with the old test. So CB is going to add some harder questions to knock down people at the top and recenter the curve to bring up the middle?
Anybody else have this impression?
Yes, the top PSAT scores are now 760 per section. I think I read that the idea is that the SAT will be a little bit harder, so an 800 PSAT only actually predicts a 760 SAT…I guess.
Overall, it looks like the manner in which hard problems are hard has changed. It used to be that problems were hardest if they required a flash of insight. Now it seems that harder problems require more training in content area – which may be what they wanted. For example, that one question where you need to find the equation of a parabola, then the equation of the line, then where they intersect…nothing inspired about it, but in terms of percentage of people who get it right, I bet that one is a 5. And there were a couple of others that gave me the same feel: harder but less interesting than old test.
Thanks for your take on things.
I can’t believe that question is a 5. You can solve it by copying the two equations into an algebra calculator. You don’t need almost any skill to do it.
I’d be interested to know what other questions you think are level 5. I couldn’t find any.
Actually, unlike the projectile one, you do have to at least come up with the equations yourself. I’ve worked with plenty of kids who could score 650 on the old test who would not have the algebra proficiency to find the equation of that parabola. (But now, effective SAT prep will include that skill.)
Seriously, I find most of the problems to be boring. Maybe that’s how to make the test less vulnerable to prep: make it too boring to focus on. Only kind of kidding here.
As I said, not much with level 5 trickiness. But if you want one: the one where they give you (ax+2)(bx+7)…it’s #15 in section 3 of the first test. I bet that stumps enough people to statistically be called a level 5.
You are right about the parabola equation. You have to know how to determine the equation of a parabola given the vertex and the y-intercept. Many students don’t know how to do that. But it is not a hard thing to do and it is really easy to teach students how to do it. You can even use your calculator to find the a of the parabola equation if you are weak in algebra. The old SAT tests things that are not so easy to teach students how to do. And after you have written the parabola equation, you can just throw the two equations in the calculator with solve. Is that a 5?
It is funny that you mention Test 1 Problem 15. I was just going over that problem with a student yesterday. I was really happy to see that although CB claims to have eliminated properties of integers from the test, the question writers snuck them back in with this question!
But ab=15 and a+b=8 is BABY properties of integers. Is that a 5? Equate coefficients is standard Math Level 1. Even some old SAT problems are equate coefficients problems. Equate coefficients is very basic. To me, the question feels like a 4. Maybe 4.5. I think the very good in math people are going to find the math too easy. Of course I could be wrong. Time will tell.
The same student (who is a junior good in reading) did the whole reading section in 32 minutes with 1 wrong. It looks like the reading is too easy too. But my sample is not statistically significant.
Here is a question from the Blue Book that is similar to Redesigned Practice Test 1 Question 15:
OG Test 3 Section 5 Question 8. But you are right. It is listed as difficulty level 5. IDK, I have seen similar questions so many times…
I found ANOTHER question like Question 15 Redesigned SAT Test 1: Question 13 from Algebra Quiz 1 in the Official Online Course.
So I guess my feeling is that maybe it was a 5 back when the Blue Book was first published, but questions like these are not going to surprise the high-scorers any more.