A stab at a cleaner, more user-friendly look at city test score data

data dump

August 10, 2011

A stab at a cleaner, more user-friendly look at city test score data

by Geoff Decker, at 1:53 pm

Click on the image to go straight to the new data below.

When the state and city education officials released the 2010-2011 ELA and Math test data on Monday, they didn’t make it easy for interested New Yorkers to make sense of the scores.

One spreadsheet, released by the city Department of Education, left off school names and corresponded results only by school code. It also excluded public charter schools entirely. The state’s spreadsheet included names, but listed every other public school in New York State as well.

There was also no easy way to compare schools to one another. The city included a comparison against previous years’ scores, but the file didn’t allow users to compare change over time among schools. The state’s data didn’t include any previous scores at all.

Not surprisingly, many of our readers emailed us to express their frustration over the scattered and unwieldy data. When I asked a DOE spokesman Matthew Mittenthal about it, he told me that grouping the data into school-by-school comparisons wasn’t a priority when publishing the information.

“We would never use test scores alone for accountability purposes, so we don’t actively encourage people to compare one school to another on that basis,” Mittenthal wrote in an email.

We spent the past couple of days playing with the spreadsheets so that it’s easier and more intuitive. First, we corresponded codes used by the DOE to actual school names (for example, 15K447 = The Math & Science Exploratory School). Then, we stripped non-essential data and added last year’s test results as a column header. Finally, we filtered the schools by performance so the best-scoring are at the top.

Here’s a first – and very preliminary – crack at sorting through the mountain of data from Monday’s release. This will be updated regularly over the next couple of days to reflect additional data-crunching, but what you’re looking at now – as of Wednesday afternoon – is a list of schools that have two years of ELA results. The current list excludes charter schools, but for each of the last two years (Note: new schools in 2010 and charter schools aren’t in this initial batch, but will be soon).

A caveat: This is a beta effort so please let us know how it can be improved and what trends you’re seeing in the comments.

Bronxmathteach

I think this also shows where the poverty is. District 7 (the south Bronx) schools are all almost entirely congregated at the bottom.
BxTeacher

I don’t see 07X162
Ellen

Thank you!
Gdecker

thanks, bxteacher – I’ll add with next update – GD
Robin

This is great. When does the state post the information on their website:

https://www.nystart.gov/publicweb/
abc123

At the very least it shows that the neediest students tend to be congregated in certain schools rather than spread out across the district/city.
http://twitter.com/BNiche B

Thank you for this valuable resource. I can see that my school that I teach at improved their number of 3/4s about 6.2%. Not necessarily a small feat, mind you.

Geoffrey (or anyone else for that matter), do you have any idea how any improvements (or lack thereof) would be reflected in a school’s report card?
District 13 parent

It’s a little obvious what the results would be, but even so I’d love to see the percent at each school that receives free lunch.

And slightly less obvious: I would expect to see that the gen-ed schools–as opposed to screened/g&t–would have a relatively even distribution of scores across the board, i.e. as many 1s as 4s, as many 2s as threes. So a gen-ed school with a typical proportion of students at the free lunch level that turned up with a majority of students scoring proficient would be beating the curve (or prepping their students far beyond the norm).
Tim

It would be a ton of work, but yeah, adding free lunch eligible, reduced-price eligible (and keeping those categories separate), and % of special ed and ELL students would make this super useful.
Tim

It’ll have a positive effect on the school’s report card–60% of the grade is based on how scores improve. But you get more points for improvement from ELLs/special ed students/minorities, if I’m not mistaken.
http://www.dianasenechal.com Diana Senechal

This is very helpful. Thank you!

What’s going on with Brighter Choice Community School (16K627)? They went up from 17.4% in 2010 to 71.8% in 2011. They only tested 39 students, but still….

A few others had unusual increases as well. Science, Technology and Research Early College High School at Erasmus (17K543) went up 21 percentage points.
RBLnyc

wow. what a great effort to put it out for actual families of students to view… thanks
A parent and a taxpayer

You have calculated a mean scale score for each school as a whole, across grades 3-8. However, the tests are not on a common scale (see http://www.p12.nysed.gov/irs/ela-math/2011/2011ScaleScoreELA.pdf), so the resulting average isn’t really very meaningful. It’s a bit like figuring out averaging an “average salary” based on salaries some of which are in U.S. dollars, some in Canadian dollars, and some in Australian dollars, for companies that have varying numbers of employees being paid in each currency. The three units of currency all have similar values–you haven’t thrown in any salaries measured in Yen, or any Regents test scores–so you end up with averages that aren’t totally meaningless, but it is still not a great practice.

If you are using the spreadsheet supplied by the state, here is how to filter out non-NYC results in Excel:
1. Turn on Auto Filtering. (Data / Fillter / Autofilter).
2. Click the button with two triangle on the right of the first column header, which should say BEDS. This will open a menu. Choose option “(Custom Filter…)”.
3. Choose BEDS “begins with” “3″. Click OK.
That should give you a list of NYC schools, both DOE and charter.

To narrow it to Manhattan, choose “begins with” “31″. For a particular district in Manhattan, choose “begins with” “31XX”, where X is the district number, with a leading 0. E.g. 03 for District 3. 32 for the Bronx, 33 for Brooklyn, 34 for Queens, 35 for SI.

For NYC charter schools, choose “begins with” “3?????86″.
To filter out charter schools, pick one of the earlier options AND “does not begin with” “3?????86″.
BxTeacher

Another website stated that Districts 12,7 and 9 were some of the lowest performing. Do you know of anyhere the data is aggregated by districts for comparison?
http://edvox.org/
new 5th grade mom

know I am dense, but, what is a mean scale score score? How is it computed for each school? thanks. new 5th grade mom
guest-manhattandad

Helpful, but skewed. Elementary schools are mixed in with middle schools and all grades are averaged in, including ‘Number Tested’ . Do you have filtering or sorting capabilities like in excel?

I downloaded the big spreadsheet and individually compared each grade and the year. Having separate rows per each school for each grade would help immensely. You could then add another column to include school average vs. school grade average. This would avoid confusion regarding differentiating scores around more important testing years (5th grade, 7th grade) vs. other years. Just some initial feedback.

GothamSchools — daily independent reporting on NYC public schools

data dump

August 10, 2011

A stab at a cleaner, more user-friendly look at city test score data

Tips, questions, feedback?

Word from Our Sponsor

Follow GothamSchools

From Our Jobs Board

Featured Employers

Recent Jobs

Chalk It Up

Recent Comments

Events Calendar

RSS Feeds

Archives

Copyright Notice

May 2013
M	T	W	T	F	S	S
« Apr
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31