I am a Sunday test-taker and I’ve known for a while that the Sunday SAT is a different test than the Saturday test. What I didn’t know is that the college board completely recycles tests. ENTIRE TESTS. Today I came home from the test and searched some of the problems on CC to find both a 2012 AND a 2014 thread that had questions identical to the test I took this morning. THEN I googled some questions to stumble upon THE ENTIRE TEST. 1. I am just completely confused, I just assumed that college board, for the most part, created new tests for each sitting. and 2. I am SO PISSEDDDDD! If I had taken this past specific test online as a practice exam, I could have gotten a perfect score. OK obviously that would be cheating and I do indeed have morals, but still??? WTH COLLEGE BOARD???
Also, any Sunday test-takers find the CR hard? I found some answer keys on this random ass Chinese website and I’m pretty sure I majorly messed up on the 5th section especially lol
Yes, the College Board has been recycling tests for years and years. It’s really unconscionable, especially since the CB is as aware as we are that (A) answers and (B) entire test scans can be found on the Internet (yes, even for unreleased tests).
I don’t get why they wouldn’t just mix up questions and produce a new test with all recycled questions, but at least they could be recycled from ** many[/b ] old tests!
You can’t just throw a lot of questions into a hat, mix them up, and then count out enough to make a test. You have to have so many level 1’s, 2’s, etc. And then one passage on science, one in fiction, two parallel structure questions, four scatterplots of different kinds, etc. Getting the right combination would be some work.
Further, the order in which a question appears affects its difficulty. The same question later in a test is “harder” (fewer students answer it correctly). The difficulty is also affected by what questions came before.
@Plotinus I think that problem could be solved pretty easily if you just divided up problems into groups (e.g. Level 1 science, Level 3 scatterplots) and picked problems for specific group types, rather than for the entire test.
@usualhopeful I don’t think any two problems have exactly the same difficulty level. CB reports them as level 2, level 3, etc., but the actual difficulty level is the percentage of students who got the question right, a number with a lot of decimal places. That means the difficulty level of a test could not be matched exactly by a set of different questions in the same areas. You would have to re-equate the remixed test.
@Plotinus True, but for that reason, no two SATs are the exact same difficulty, and I think it would actually be easier to create a test of appropriate difficulty because the questions would already be tested and have percentiles associated with them.
@usualhopeful Your idea is solid. But, it would require more work for CB, so they’ll never adopt it. And @Plotinus: No two tests will ever have the EXACT SAME difficulty level when you out to ten decimals.
Even still, it’s wrong that CB reuses problems in the first place. Certain problems are ‘memorable’ and hence the answers are easily remembered. What I mean is that no one needs to memorize the easy math questions, etc. However, for the 1-2 challenging (aka tricky) math questions, it’d be nice to have some muscle memory (i.e. the answer to the Johnson Farm question is 18 legs).
No two tests have the same difficulty=CB has to come up with a different scoring curve for the remixed test to compensate for the different difficulty level. This is equating. It takes time, money, and sufficient data.
I’d rather see kids taking tests of slightly different difficulty than the widespread shameless cheating which is going on with the current system. That’s completely unfair to the kids who aren’t sleazy cheaters and it calls the entire system into questions. Also, much as the college board would like us to believe these SAT scores are handed down from God, actually, we see lots of kids taking and retaking and the scores bounce up and down quite a lot. And at the higher levels the wonderful equating process just ends up meaning if you got a few wrong on one day you are getting a lower score than if you got the same number wrong on another day. The fact that kids with scores hundreds of points lower than you may have found the problems more or less difficult often isn’t relevant to the high performing students. The issues with mixing up problems from old tests are really quite minor in the context of these greater problems.
@Plotinus But as it is now, no two tests have the exact same difficulty because the CB is writing new questions and doesn’t know exactly what curve they will produce.
The redesigned SAT is a very different test from the old SAT. The test forms of completely different tests are not equated but concorded. This is a different process. I haven’t seen any Truth-in-Testing information about what concordance process CB will use. If it is anything like the ACT-SAT concordance, then on average a person who scored x on the old SAT will score y=the concorded equivalent on the new SAT. However, as in the case of the ACT, some people will do better or worse on the new SAT than the concorded equivalent of their old SAT scores. It will depend upon whether you have more or less of the specific skill set tested by the new SAT than does the average test-taker. There is a lot of room here for data fudging.
I doubt any of this will influence CB’s recycling practices. My guess is that CB will just recycle test forms of the new SAT the same way it recycled test forms of the old SAT, especially since there are going to be even more administrations (including SAT School Day in the Spring and, starting 2017, an administration in August).
The CB is planning on only releasing a few tests a year, as it has in the past. This clearly suggests that it will continue its recycling practice and that cheating will continue unabated as a result.