4 ms·

> But we're already measuring that variance directly. >> To add to this, they need not be wrong (and really should not be to make the argument), just significa

> But we're already measuring that variance directly. >> To add to this, they need not be wrong (and really should not be to make the argument), just significantly incomplete. From what I gathered from this thread, this was already agreed upon by everyone. ...and... >>> Unless those tests are NOT measuring performance in a meaningful way. ...and... >>>> Otherwise, it's just weighing one questionable feature more than another questionable feature. ...and... >>>>(or something)>>> I think the fundamental thesis of the article -- that we can scientifically manage a system of universal education -- is fundamentally and impossibly flawed. Especially given the current state of assessments and the incentive structures built around them. And so on... So there's a general agreement that the current tests (and/or the social system they're embedded in) are extremely flawed and at least a contest over assertions about whether good enough tests event exist. Which is to say, for the fifth(?) time, the basic problem is that standardized tests are not doing even a remotely reasonable job at "measuring that variance directly". So pretending like they are legitimate metrics (and/or any more legitimate than other fairly broken metrics like funding levels) for measuring overall educational attainment is at best flawed and at worst -- I assert very likely! -- the very source of the whole problem with K12 in this country. On that note, I would like to close out this thread by imploring you to read the essay I linked to in my first post: https://www.maa.org/external_archive/devlin/LockhartsLament.pdf https://www.maa.org/external_archive/devlin/LockhartsLament....

darawk 8y ago
You didn't respond to my point at all. There is a true signal, and there is the measured signal. We're measuring, obviously, the measured signal. It is, according to you, a weak proxy for the true signal. That is, it correlates with it at some level. Secondly, we have another variable 'spending'. We can measure that this 'spending' variable correlates with the measured variable. Therefore, we know that those two variables are related in some way. Now, in terms of getting a bead on the true signal, do we care about the spending variable? The only reason we have to believe it has any relationship to the true signal is that it correlates with the measured signal! If we didn't know the measured signal directly, that might be useful. But we do. What I am saying is this: The performance tests are the best we have. There is no additive value of this 'spending' variable, as far as anyone i'm aware of can tell. If you have information to the contrary, i'm curious to hear it, though. Now, you may, perhaps even rightly, that say that the performance tests we have are so bad that they're useless. Well, i'm not going to take a position on that. The position that i'm taking is that adding 'spending' to your model does not minimize your loss function here, and there is no reason to believe that it would.
- thrmsforbfast 8y ago
  > If you have information to the contrary, i'm curious to hear it, though. Providing such evidence was literally the main claim made in the top-level comment on this now very long thread... Also, the fact that spending correlates well with other test-based performance metrics (see: the top-level comment on this now very long thread). And that the state rankings that result from the NAEP-based ranking are wildly inconsistent if you construct rankings based on other tests (e.g., state exams, college extract exam scores).
  - darawk 8y ago
    > Providing such evidence was literally the main claim made in the top-level comment on this now very long thread... You provided evidence that spending has additive value in predicting outcomes? Please, cite it again. > Also, the fact that spending correlates well with other test-based performance metrics (see: the top-level comment on this now very long thread). You want to use, as evidence to support the view that spending provides novel information over and above performance evaluation, that it correlates with performance evaluation?

darawk 8y ago

You didn't respond to my point at all. There is a true signal, and there is the measured signal. We're measuring, obviously, the measured signal. It is, according to you, a weak proxy for the true signal. That is, it correlates with it at some level. Secondly, we have another variable 'spending'. We can measure that this 'spending' variable correlates with the measured variable. Therefore, we know that those two variables are related in some way. Now, in terms of getting a bead on the true signal, do we care about the spending variable? The only reason we have to believe it has any relationship to the true signal is that it correlates with the measured signal! If we didn't know the measured signal directly, that might be useful. But we do. What I am saying is this: The performance tests are the best we have. There is no additive value of this 'spending' variable, as far as anyone i'm aware of can tell. If you have information to the contrary, i'm curious to hear it, though. Now, you may, perhaps even rightly, that say that the performance tests we have are so bad that they're useless. Well, i'm not going to take a position on that. The position that i'm taking is that adding 'spending' to your model does not minimize your loss function here, and there is no reason to believe that it would.

thrmsforbfast 8y ago
> If you have information to the contrary, i'm curious to hear it, though. Providing such evidence was literally the main claim made in the top-level comment on this now very long thread... Also, the fact that spending correlates well with other test-based performance metrics (see: the top-level comment on this now very long thread). And that the state rankings that result from the NAEP-based ranking are wildly inconsistent if you construct rankings based on other tests (e.g., state exams, college extract exam scores).
- darawk 8y ago
  > Providing such evidence was literally the main claim made in the top-level comment on this now very long thread... You provided evidence that spending has additive value in predicting outcomes? Please, cite it again. > Also, the fact that spending correlates well with other test-based performance metrics (see: the top-level comment on this now very long thread). You want to use, as evidence to support the view that spending provides novel information over and above performance evaluation, that it correlates with performance evaluation?