I’m curious what their underlying technique is. It seems reasonable to simply look at the number of correct answers by each designated group (race, gender, etc.) and single out the questions with widely varying distributions. I think it would be…ummm…better? interesting?..if they instead added questions which adjusted the bias. The technique to do this would be extremely difficult, but I think it would make a fairer test.
For instance, the concept of “dune is to sand as drift is to snow” would have a similar corrolary in the south with perhaps “grits is to corn…”, or sticking to meteorology, “Beaufort is to hurricane…” Questions in this section are designed to test reading level, those who are “well-read” should excel over those who simply have regional or cultural knowledge.
…if they instead added questions which adjusted the bias [by group]
NeedAHobby
The idea is to ELIMINATE bias down to each individual question, so that the overall test become unbiased. By deliberately biasing by group so that the “average” result of the whole is statistically unbiased is really to exacerbate the situation. It will also call for everyone who wishes a 1600 to have to perhaps live and study every single biased culture that exists. While this would be commendable as an educational goal, it is social engineering in the ultimate. It is calling for separate and individual tests by social groups, which really defeats the purpose of a “national” standard by which everyone can live by.
For instance, the concept of “dune is to sand as drift is to snow” would have a similar
corrolary in the south with perhaps “grits is to corn…”, or sticking to meteorology,
“Beaufort is to hurricane…”
This is not the method that ETS uses, and for good reason. I think we all (Americans, I mean)would have a bit to complain about if the test to get into Oxford was full of qustions about blood pudding, rugby, and British parlimentary government.
I sure don’t want questions about grits affecting my chances to get into college, and lived in southeast Georgia for two years. I still don’t know what they really are.
As a sort of aside, if you have a relative or friend that is trying to study for the SAT, I highly recommend that they buy the book “Ten Real SATs.” All the other books out there have questions that would never be used on a real test. Yes, all of them. My day-to-day job is teaching kids the SAT, and I can state catagorically that I have yet to find a book by another company that does not contain questions with arguable answers, or even worse, like the Gruber’s book I was looking at yesterday, questions about things like baseball.
Run:baseball::
The answer is supposed to be goal:soccer.
You explain to some bright Taiwanese girl why she can’t get into American college because she doesn’t know that a point in baseball is called a run. Absolutely a terrible question. They should be ashamed to have it in their book. It just goes to show you that there’s no incentive to get something right like being sued for racial or cultural bias, and none of these book writers ever are.
Hmmm…I’m wrestling with the concept of “better for the test” and “better for society.” Obviously the SATs are not the be-all end-all of society, but I think it’s important to be well-rounded (in the sense of being open minded) when entering college.
I guess it all depends on the college. Perhaps a cultural awareness test could be created to see how diverse someone’s mindset is. The SATs can strive to be unbiased (albeit using English, an inherently human-biased language–i.e when did “female dog” slip from definition #1 to #2 for the word bitch, at least in common usage?).
Of course, this all sounds like it would be accomplished better in the vein of the Purity Test–sort of unofficial and dynamic. Sounds like a plan…something else to put on the great list.
