A NY Times article discusses a curious finding across the country: when school systems implement a teacher evaluation system, almost all the teachers receive good ratings. Given the conventional wisdom about the decline in US public education, finding that most teachers are quite competent would indeed be curious to most.
Of course, it depends on the threshold for competency. Evidence suggests this threshold is quite low:
The changes, already under way in some cities and states, are intended to provide meaningful feedback and, critically, to weed out weak performers. And here are some of the early results:
In Florida, 97 percent of teachers were deemed effective or highly effective in the most recent evaluations. In Tennessee, 98 percent of teachers were judged to be “at expectations.”
In Michigan, 98 percent of teachers were rated effective or better.
Advocates of education reform concede that such rosy numbers, after many millions of dollars developing the new systems and thousands of hours of training, are worrisome.
They should be worried. The question is one of incentives and shared interest. Teacher unions are against evaluation in part because it partially unravels the pact that unions make with their constituents: namely, that in exchange for finance and obedience, the union will organize a united front to ensure that there are no losers within the group. The consequences of this contract are job security for members with the trade off of wage compression where high performers receive less and poor performers receive more to maintain parity. In extreme cases, the only thing that matters for wages is seniority and possession of master’s degree. Hence, evaluations help to unravel this pact because if they are accurate, they have the potential to cause dissension within the ranks especially if they are paired with merit pay.
Further, the switch to formal evaluations has been fraught with problems in part because evaluations help set up adversarial relationships between administrators (previous teachers) and teachers.
So how do we avoid all these problems? Set the thresholds so low that all but the most inept will pass. Thus, you reinforce parity among teachers, don’t screw over your friends, and don’t piss off the unions.
Its not clear what is the best design, but the current one is not working.
As a native Chicagoan and, after some time away, a current resident, I have to take issue with the way Chicago has been recently characterized by certain media outlets and pundits. While murders are still quite high and we have way too many tragedies where innocents have been taken, Chicago is a great place to live. Even the Southside is not as Kevin Williamson describes it, a “wasteland,” below 35th St, although some areas sure look like it.
Of course, this is not to say that poor people are not catching it here and that more could be done. It is easy for me to say that it’s not that bad as I am fairly privileged now; my kids have little worry in my current neighborhood. Nevertheless, by characterizing Chicago as a “war zone,” one diminishes the plight of the poor in the city and poor suburbs and especially those people who live in actual war zones.
Chris Blattman links to a recent NBER working paper by Card and Dellavigna by half-seriously noting that his anxiety levels (about tenure) are relatively high:
The first four facts:
- Annual submissions to the top-5 journals nearly doubled from 1990 to 2012.
- The total number of articles published in these journals actually declined from 400 per year in the late 1970s to 300 per year most recently. As a result, the acceptance rate has fallen from 15% to 6%, with potential implications for the career progression of young scholars.
- One journal, the American Economic Review, now accounts for 40% of top-5 publications, up from 25% in the 1970s.
- Recently published papers are on average 3 times longer than they were in the 1970s, contributing to the relative shortage of journal space.
I’m even more anxious than he is.
- R: You are willing to invest in learning something difficult. You do not care about aesthetics, only availability of packages and getting results quickly.
- Python or JVM languages: You are a hacker who may have already been a programmer before you delved into statistics. You are probably willing to run alpha or beta-quality algorithms because the statistical package ecosystem is still evolving. You care about integrating your statistics code into a production codebase.
- Julia: You are John Myles White.
- Stata: You are an economist who doesn’t care to code your own estimators, probably because your comparative advantage lies elsewhere. Possibly you are doing sophisticated work with panel data where Stata is the only game in town. You don’t care that you can’t do proper programming because you’re not a programmer.
- SPSS: You love using your mouse and discovering options using menus. You are nervous about writing code and probably manage your data in Microsoft Excel.
- Matlab: You definitely know what you’re doing and you care about performance. You know Matlab is expensive but you aren’t the one paying for it. You live in a bubble where everyone you know uses Matlab.
- Mathematica: You are an aesthete who believes everything Stephen Wolfram says.
SAS: You are an analyst for a large pharmaceutical company, and SAS is all you have ever known. You have a large library of custom SAS macros, so that (clearly) makes you a programmer. That anyone would want to hand-code statistical methods leaves you utterly baffled. If SAS does not ship with a particular statistical method, then it probably isn’t important. (h/t Chris Fonnesbeck)
Snarky and judgmental, eh?
I’m not sure what to take away from this. While I agree that the availability of canned point and click type statistical software can lead to problematic analyses in the hands of the untrained, I wouldn’t assume one doesn’t know their stuff because they use a canned software package. Some of the most famous and well-known econometricians use Stata for testing out proposed estimators. By contrast, I’ve seen individuals who don’t understand what a parameter represents, use R and Matlab to do analyses.
Depending on the task I use a variety of tools in my analyses: Python (Numpy, Scipy) with R (via Rpy2), Stata, Fortran, and Matlab. It depends on what I need to do and the most efficient way to do it. What’s my signal? I’m unable to commit?
If it is a signal, it is not a very informative one without additional information.
With all the discussion on mass gun violence and mental illness in the media, I think it is worthwhile to remember that if even one life is lost due to violence or mental illness, we should all be concerned. Consider the recent “hate crime” in NY. An Indian Hindu man named Sunando Sen was pushed to his death by a woman who ostensibly suffered from some mental malady. Her reasons:
“I pushed a Muslim off the train tracks because I hate Hindus and Muslims ever since 2001 when they put down the twin towers I’ve been beating them up.”
Of course, there is no evidence that Hindus participated in any part of the 9/11 terrorist attacks. I guess in her mind if they look alike, they must be the same. Of course, our Sikh friends have been experiencing this type of negative treatment for wearing turbans for some time now.
While it was not 5 or 10 people, the pain of this man’s family is the same as any one person who was killed in a mass attack. And this killing required no gun, just an opportunity.