The Test is "Quirked"

[rabow]

The NYT reports on a scoring quirk of the admissions exam.

Gothamist points out the following quote:
According to American Guidance Service, 52 students - or one-fifth of 1 percent of the more than 25,000 eighth graders who took the test last year - scored perfectly on one part. Of these students, one scored in the 50 percentile on the other part

Comments

[miserablexone]

Man, that's weird. I think I probably did better on the verbal section. I totes pwned the scrambled paragraphs. I wish they had those on the SATs.

[corneredangel]

Trying to say that the test has "issues" isn't really right, though, since at least from the article, the scoring *is* actually designed to reward someone who is perfect in one thing and all right in the other.

I wonder what the overall distribution of scores is, though, and where the average score lies. It's got to be very low, for only what eight hundred people to get in of over twenty-five thousand taking the test.

[rabow]

Text is changed.

The article's tone makes it pretty clear that the NYT thinks there are issues with the way the test is scored. Although as apotwixx notes, most of us don't really mind...

I think it becomes more of an issue with the schools that have been tacked on past Brooklyn Tech. Even with the new schools, though, I think they're letting less than 50% of people who take the test get in. (30%? 40%?) A current test prep book probably has those cutoffs...

[corneredangel]

I think it becomes more of an issue with the schools that have been tacked on past Brooklyn Tech. Even with the new schools, though, I think they're letting less than 50% of people who take the test get in.

Vastly less, I'd say. What, twenty-five thousand people take the test, for maybe five thousand spaces total, all told.

Mostly what annoys me about the article is that they barely mention that it's not privileding *math* per se, but rather, either side of it.

[apotwixx]

I can remember going to old Stuy to take that test back in 1989. I think the building was more imposing than the test. I remember I made the cutoff for Stuy by 20+ points. My scores if I remember correctly were pretty balanced. Hmmm... this does make you think a bit. Overall I think the test does a good job based on the classmates I had. :-)

[mokey4]

How did you know how balanced your test scores were? From what I recall (and I took the test around the same time as you), we were only told our total score on a 200-800 scale, and not a math/verbal breakdown.

I barely got into the school (I think I might have beat the cutoff by 2 points if I was lucky), and I know I was somewhere in the bottom half of my graduating class. But as Ms. Schmidt famously said to my mother on open school night, "I like Danielle because she doesn't care about her grades, unlike most Stuy students she's really in it for the learning." That wasn't exactly true- I liked learning but I was too lazy to go for the high grades. That hasn't exactly changed.

[mephistakitten]

I think that the test is not infalliable for predicting who is going to have a successful academic career and who isn't... case in point...

I got in to Stuy by 3 points, but ended up in the top 75-80% of the class.

[larrytc]

Doesn't that mean you just beat 20-25% of the class? =P

"Stuyvesant loves lopsided geniuses."

[_darkvictory]

Love this quote from the math education consultant.

Re:Lopsided Quotes

[rabow]

Another great quote is from the test-prep guy (a Stuy grad himself!) who doesn't focus on the scoring method. How can you prep students for tests without focusing on the test's scoring?

Granted, the scoring shouldn't be emphasized as much as content, but it should at least be a factor...