3 ms·
The LA times had an interesting series (http://articles.latimes.com/2010/aug/22/local/la-me-schools-about-20100822 http://articles.latimes.com/2010/aug/22/local
by ylem 16y ago
The LA times had an interesting series (http://articles.latimes.com/2010/aug/22/local/la-me-schools-about-20100822 http://articles.latimes.com/2010/aug/22/local/la-me-schools-...) on value added metrics in education that was interesting. The basic idea is that you look at the median achievement scores for students the year before they enter a teacher's classroom and the year after and see how they compare (of course this is vulnerable to schools with students who transfer frequently). Then, you can say for example, on average does a student who was below grade level, stay below grade level, or improve? This allows one to get some handle on what the impact is of the teacher--for good or worse--the LAtimes study found that the differences in teacher impact were rather high--within the same school. Also, some teachers were happy to learn about the results--for example, some teachers were enthusiastic, but were not aware that their overall impact was not so high and wanted to know how they could improve. Now, I wouldn't use this as the only metric because it might not be linear--for example, perhaps it's harder to get a large improvement for say a group of gifted kids (relative to their previous scores), compared to moving students who are somewhat below grade level to slightly above grade level--but at least it indicates the direction that the teacher is (on average) sending students. I should stress that the method looks at the same student before and after and then averages the results afterwards.
- andrewce 16y agoI'm okay with value-added metrics, so long as it is not the only tool used to evaluate and assess. 1) Not all things taught appear on a test, or can be easily measured, and 2) Teachers are not the only entities responsible for student growth. There is also the curriculum, the social environment, the home life, and so on. 3) The natural reaction to such metrics, particularly in the current educational climate, is "Holy crap. If I don't do well on this, I'll get fired." Rather than jumping on the "Let's fire 'em" wagon, I feel that comprehensive professional development plans, based around highlighting strengths and excellent patterns, would accomplish just as much. Instead, though, the current professional development plan seems to be "Fire 'em".
- yummyfajitas 16y agoMost of what you describe is not a problem, due to the law of large numbers. Regarding (1), it's very true that a test is just a sample of the material taught. Say you teach 90% of the material well, 10% poorly. The probability of the test being comprised primarily of the 10% you taught poorly is highly improbable. Regarding (2), this is irrelevant as well. VAM compares your group of 30 low income students in district A being taught curriculum X to the overall average of low income students in district A being taught curriculum X. While it's possible that all your 90-150 students (depending on how many classes you teach) have a worse home life than the average low income student in district A, the probability of that is very low.