Posted on 2010-08-27 12:46:00
I've been accumulating these links, and then I kinda forgot about them, so here goes!
The LA Times just did a big investigation on teacher performance. I fully support using data like this (they use a "value-added" approach, where they compare students performance at the beginning and end of the year) as part of teach evaluations. Surely people can study what the good teachers are doing right and use that to help all teachers.
Comparing the tax plans - the graphic is a little confusing. What surprised me is that Obama's plan (which is to keep the Bush tax cuts for incomes < $250K and index the AMT to inflation) costs $3 trillion over 10 years, while just keeping all of the Bush tax cuts only costs $.7 trillion more.
In last week's Futurama episode, one of the writers (who has a PhD in math) proved a theorem that was instrumental to the plot - they even showed the proof briefly!
Did you know: the Blue Power Ranger quit the show because he was harassed for being gay by the producers, etc. In other news, Ken Mehlman (who ran Bush's campaign in 2004 and was head of the RNC) came out and is now supporting the legal challenge to California's Prop 8.
The evolutionary case against monogamy - I'm in the middle of reviewing their book. It is very interesting.
Comment from skimmerduk:
I think that analyzing teacher impact on student achievement is a really good thing that is undoubtedly painful but also essential for effective education. If teachers are allowed to attribute poor student performance to five billion external factors over which they have no control, then it ultimately makes their job seem futile. Teachers need to believe that their instructional efforts do have consequences, and really the only way to internalize that is to see it in some concrete, verifiable way.
I do have a question about the value-added approach in this article, however. Since the percentiles used for comparison seemed to be based on district norms, doesn't that mean that, for one teacher's class to improve, another teacher's class must inevitably do worse? This seems like it would give you a way to rank teachers across the district in comparison to each other, but if all the teachers were actually effective (i.e. all their students were learning something), it wouldn't actually indicate that anyone was doing a BAD job; just not as good a job as someone else. I've obviously never been too comfortable wrapping my head around these kinds of things, so this really is a question rather than a challenge.
Comment from gregstoll:
That is a good point (and one I thought of when reading the article and then forgot). Percentile is an OK method at first, but some sort of absolute metric (like score on the tests or something) would seem to be better.
Comment from djedi:
I agree. As a student, I always disliked how much I was judged not on what I knew but how I performed relative to others.
This backup was done by LJBackup.