HAYAO
Science Advisor
Gold Member
- 379
- 238
PeterDonis said:Not necessarily "wrong", just that "good test taker" is not the same as what is actually desired: a student who can put the knowledge learned to actual use.
Which is much, much harder than your remark here suggests.
Yeah, exactly, there's no doubt about it. Have I denied that anywhere in my posts?PeterDonis said:None of this addresses my concern. My concern is not that "good test taker" has zero correlation with things we actually care about; I never made any such claim. My concern is that "good test taker" is not the same as the things we actually care about. The things we actually care about probably cannot be tested directly in an artificial environment at all. Whether or not standardized tests, or grades in school, are a good enough proxy for the things we actually care about to make using them as metrics worthwhile is a difficult question involving a lot of judgment calls; it is not the straightforward cookie cutter process you appear to believe it is.
Look, we actually agree on most points, and we are drifting away from what actually needs to be discussed. I am merely stating that the tests we have today is the best we've got right now. With the most rigorously general one being the IQ test. Standardized tests as Dale pointed out are also feasible approach. So we have to work with what we have. Teachers (ideally) put hours and hours of effort in making tests that hopefully assess what it's supposed to test, so that's the best we've got. You can't just say "well the test don't account for this and that". What do you suggest we do?
I believe literally everyone here understand how difficult it is to make a "good test". Stop assuming that teachers don't know that. I teach in college, I have my license for grade-school as well. I know what I'm talking about.
Also, I never claimed that you think a "good test taker" has zero correlation. Where have I said that?
The statement about IQ tests addresses your point exactly. You think that tests can't distinguish between "real thinkers" and "good test takers". That's true, but the point is, what do you suggest we do? IQ test is one of the most rigorous general cognitive ability test developed by science, despite not being perfect. People criticize IQ tests for their inadequacy, but it's still the best we've got. The 'best we got' has an correlation with career success up to merely 0.4. I want you to put that into perspective. You are making a pointless statement when you talk about "grades don't fully reflect this and that" because everyone knows that. Yeah sure. In that case, PhD degrees also don't reflect the actual academic ability of the holder but is almost always the prerequisite to get a position in college. So, should PhD be abandoned concept? Or should be replaced with something else? Can you suggest an alternative? The answer is, we can't do anything about the test further than the best we have today. But it's still better than most alternatives, so we have to use it anyway.
At least in my opinion, the alternative is definitely not using homework as a way to grade. (In case you might try to twist what I said, I am not saying you claimed that we should use homework to grade.)
As a side note, I always recommend combination of projects and tests for classes and grade solely based on those two. Tests tend to test one's ability within a given time limit to answer specific questions. Project on the other hand have some level of abstraction that students can freely experiment with. I also suggest teachers use combination of specific problem based projects and fully abstract project (students come up with their own question). I also suggest using standardized tests in combination with normal tests that teachers create because there's some specificity of those test and what it assess.
Last edited: