Arnold Kling wondered aloud what the relationship would be between measures of a school district’s poverty level and performance on standardized Math exams.
I spent some time this morning putting together 2005 data for New York State. New York has 698 school districts, but the plot below includes data only for 628 because of data issues (not reporting of districts, or no students taking a math regents … is that even possible?). The data come from 2005 because all of the data is not available yet for more recent years, and even then, I am not any good at extracting Access data (older data is in Excel). The chart below plots on the X-axis (the horizontal one) a measure of the poverty level in the school district (the New York State Education Department calculates an index for every district) and on the Y-axis plots the percentage of students passing a Math regents exam in the 10th grade.
The yellow line indicates a simple regression line through the data points. It says that school districts that have less poverty have a higher share of students passing the Math regents. This is completely unsurprising (the equation is in the box). Schools that are represented in dots above the line are those that have a higher share of students passing the regents than would be predicted by the level of poverty in their district (again, I warn that this is a simple regression controlling for nothing else), and those represented by dots below the regression line are those that have a smaller share of students pass the Math Regents than would be predicted by their level of poverty.
In summary, those districts above the line are over-performing while those below are under-performing. But the interesting question Arnold poses is how do expenditures per student compare for those districts that are over-performing versus those that are underperforming? Is the better performance because those schools have more money? You judge for yourself:
- Districts above the line spent, on average in 2005, $14,047 per student;
- Districts below the line spent, on average in 2005, $13,979 per student;
The over-performers outspent the under-performers by about the cost of a social studies textbook.
Michael,
The data appear to be highly nonlinear. I suggest you use decile groupings to rescale the data. That is, a school district gets a 10 if it’s in the top decile for pass rate, 9, if it’s in the next decile, etc. Do the same with the inverse poverty index. Then run the regression and see what happens.
One thing to keep in mind are cost of living differences, which are substantial between, say, Manhattan and Cooperstown. Manhattan schools have to pay teachers far more than small town schools.
You can get cost of living estimates for all small cities on up from ACCRA.
Sign: wdpad Hello!!! uhvfw and 5357zkxpdaumpk and 1460 : I love your site. 🙂 Love design!!! I just came across your blog and wanted to say that Ive really enjoyed browsing your blog posts.
Sign: zdbrw Hello!!! wsusb and 2410vbwkdrfzrf and 4653 : Cooooool blog really
Bet on UFC, US players are welcome, recieve 140% more from first deposit ! bet on ufc
Strikeforce odds
709231 137861Correct wow messages are bound to show your and supply memorialize the speacial couple. Beginner sound system to high in volume crowds really should always take a appear at all with the amazing value behind presenting and public speaking, which is to be someones truck. best man speeches brother 950999
266088 384586There is noticeably a bundle comprehend this. I suppose you made specific good points in functions also. 528387
554309 46114Extremely clean web site , thanks for this post. 989080
926018 815641What web host are you the use of? Can I am getting affiliate hyperlink in your host? I wish web site loaded up as quickly as yours lol 556800
121851 205376Having been basically seeking at valuable blog articles with regard towards the project research when My partner and i happened to stumble on yours. Thanks for this practical information! 178148
133784 871095So may be the green tea i buy in cans exactly the same as the regular tea youd buy to put inside your morning cup? Ive been told is just normal green tea produced to be cooler, but does it have any affect as far as not speeding up your metabolism as quickly as normal hot green tea? 995930
831295 366937Youre so cool! I dont suppose Ive learn something like this before. So nice to search out any person with some special thoughts on this topic. realy thank you for starting this up. this internet internet site is one thing thats needed on the net, someone with a bit of originality. beneficial job for bringing something new towards the internet! 982462