14 ms·
This study seems to ignore the most obvious differences between standardized tests and typical classroom grades. Standardized tests tend to capture a single typ
by skywhopper 4y ago
This study seems to ignore the most obvious differences between standardized tests and typical classroom grades. Standardized tests tend to capture a single type of work (getting the right answer, regardless of process) in a single moment (test day) with an objective evaluation (multiple choice). Whereas classroom grades take into account various sorts of work, including process work, over time, and inevitably take into account a variety of skills and performance beyond the “right answer on test day”.
Which is “better”? Impossible to say. Neither captures the whole picture. But the headline and apparent conclusion of the study is misleading at best.
- algem 4y agoin the article they gave the same standard test to teachers to mark and it showed the exact same bias as other tests.
- PaulHoule 4y agoOne argument for standardized tests is that the classroom allows people with high social capital but low academic ability many ways to bullshit their way to higher grades. Taking practice tests and familiarization with standardized tests helps, but a person who can't test their way out of a paper bag can do very little to improve their standardized test scores other than hire a ringer to take the test for them.
- gragundier 4y agoI think a mixture is probably the best, because, cynically, I bet people with high social capital could engineer a way to improve their standardized grades as well. By requiring them to cover more bases, it drives their costs up. A ringer is an obvious solution, but there is a lot of room to be "creative." The best way for admissions to pick well is to always remain liquid on metrics. Keep the pool wide; diversify your metrics portfolio.