It is horrible how much of a disservice “well meaning” people are doing to students.
Get a 35/100 on a state exam and pass.Get a 99 on the same exam and get your score lowered.Insanity.
It makes sense to me. The assigned score for the test is not the percent correct, just like the assigned score for an SAT is not the percent correct. And I can easily see where getting 1/3 of the answers right on a difficult test is enough to pass: there are tests where getting 1/3 of the answers right is a sparkling score. I can’t say whether this test is so difficult that getting 1/3 of the answers right should be enough to pass, but in principle, low scores on hard tests are better than high scores on easy tests.
This is exactly right. I really think this has nothing to do with common core, and it annoys me to read propaganda designed to push people’s buttons (BTW - I’m neither a proponent nor a defender of common core; selfishly, I really don’t care about it because my kids are too old for it to affect them.)
Respectfully, the algebra teacher who’s complaining in the linked article has no idea what she’s talking about when it comes to rigorous test design (not her fault for not knowing; most teachers don’t know much about it). If you want to have a test that accurately differentiates ability levels with high validity etc. then you want your test to have a good mix of easy/medium/hard questions and to result in a wide range of raw score distributions. That’s why it’s actually good to have a wide range of raw scores, in this case say from 30/86 to 85/86 (except for the effects on student morale !).
Statistically speaking, if someone wants to complain about anything they should complain about people’s egos being so fragile that we still have to make a score of ‘F’ (failure to understand the rudiments of algebra) equal a equated score that’s as high as 64%
There are statistical specialists (called “psychometricians”) who figure out how to do equating in a rigorous way. The simple “add 10% to everyone’s score” that many teachers do (especially in high school) completely breaks down if you are testing students who have wide range of ability levels and want to be able to scientifically compare an SAT score of 600 earned in the year 2000 to one earned in the year 2015 by a different generation of students.
Contrast that with the “normal” high school algebra curve that most teachers use. They probably want 75% of their students to get at least a B- = 80%. This creates 2 bad problems - (1) it means you probably need to have the bulk of the test be easy enough that almost any student with a heartbeat can get 65% of the points almost for “free”, and (2) all the good performers are scrunched into a range between 93% and 100%, so the test tends to reward obsessive attention to minute details rather than mastery and sophisticated understanding of the advanced material.
That being said, I don’t know how sophisticated the NY Regents people are. I believe they have an excellent reputation though, and if I had to bet I’d say they’re doing things right.
You also want the good students to get the hard questions wrong and the easy questions right. i.e. a questions that has everyone guessing isn’t a good question. When dh was designing questions for histology they worried a lot about this, and ended up throwing out questions that didn’t conform to the curve.
My kids didn’t think the Regents tests were hard, but there were some years when the tests were unusually difficult and there was a lot of flak about it. (The chemistry test the year before my oldest took chem notably ended up with a lot of people getting rescored because they threw out so many questions.)
The old Regents algebra test (pre Common Core) had similar curves and that test was not difficult.
The curves are not similar. A raw score scaled to a 70 on the January, 2015 CC Algebra test would have yielded about an 82 on the Integrated Algebra exam.
Ok @Mehitabelle I didn’t compare the entire curves but the low points are similar. In 2013, a 30 out of 87 was a “65” on the integrated regents and a 70 out of 87 is a “86”. http://www.nysedregents.org/integratedalgebra/113/ialg12013-cc.pdf
The linked article (OP) says a 30 out of 86 is a 65 this year.
Isn’t this the system used on most standardized tests? Seems like a bunch of hoopla about nothing.
The wonder of standardized tests.
That algebra teacher needs to take some more advanced math courses.
There is a problem with the NY State Algebra 1 Common Core Regent exams. It’s true that conversions of scores from % to a standardized score that might be a number lower than the % correct on the test is not necessarily problematic. And it is true that student and teachers often do not understand psychometric properties of tests. But, a related problem is that the score for the Regents is often averaged into the student’s grade-sometimes it is worth as much as a quarter’s grade. And then, the premise is that the scores are on a ratio scale-that the distance from 75 to 76 is the same as the distance from 89 to 90. but that assumption is violated. Of course grades are so messed up psychometrically speaking anyway. But, this adds a new wrinkle that is obvious and problematic.
The curve is designed to meet the political agenda.
I have a better idea, just take all the scores and average them and give every student an average grade. No one would shine but I’m pretty sure most would pass.
All the way through high school our passing grade was 40%. 70% was an A. Teachers gave 80% once in a blue moon, if ever. A system designed so that 93% is an A, and most kids get an A, is making it too easy.
35% or any other passing grade is reasonable or not depending on the stringency of the test, not the absolute score.