This is a map of champaign-urbana schools (Champaign Schools.pdf), color coded by school type (public/private) and 2004-2005 performance for public schools, against a backdrop of 2000 census block by percentage of minority residents in CU. This is in response to the thread: http://www.illinipundit.com/2006/10/16/charter-school-for-champaign-boys#comment and http://www.ucimc.org/node/497.
The school layer does not differentiate what level of schooling (k-5, middle, jr, high).
It does, however, indicate whether or not the school made AYP in 2005, if the school is a “spotlight” school or an “AEWS” (Academic Early Warning) school per the interactive schools report from NIU, or no indication for either (“OK”).
Non-spatial data for the public schools came from: http://iirc.niu.edu/.
Spatial data (street addresses) for all schools came from: http://Illinois.publicschoolsreport.com and http://Illinois.privateschoolsreport.com. Street addresses for all schools were geocoded to a street shapefile from the Champaign County Regional Planning Commission (CCGISC), current as of Oct ‘06. Minor local streets were left out for clarity, major roads were left in for visual spatial reference.
2000 demographic data (spatial and non-spatial), water features and rail features came from US Census.
Minority percentage was derived for each census block by creating a field for “(POP2000 - POP_WHITE) / POP2000,” and then equal-break visualizing the blocks in 5% intervals, with darker shades indicating a higher percentage of minorities in that block.
Census blocks with no color indicate that there is a 2000 population of “0” in that block.
I have no idea of the spatial makeup of each school (ie, what percentage of students come from what part of town, or demographic makeup of each school), so no correlation should be inferred between location of a school, performance of a school, and demographic breakdown within each school.
There is no intentional political motive behind the map; it is designed to be an unbiased point of reference for the blog users to work with, so we’re all on the same sheet of music.
UDPATE BY IP: The file is here. I'm not sure why anonymous users couldn't see it before , but I'm trying to fix it.
UPDATE BY IP #2: Fixed now. You can view the file just below the post, whether you're anonymous or not. I love Drupal!
| Attachment | Size |
|---|---|
| Champaign Schools.pdf | 329.91 KB |
| Champaign Density.pdf | 328.79 KB |
| Champaign Parcels Schools.pdf | 796.51 KB |
I promoted this to the front page, as this is exactly the type of user-generated content that I think the community needs and wants to see.
so where is the map?
You have to register and log in to the site in order to get to the link to download the map.
That's a nice map - the color scheme works, and using census blocks as units was a good choice.
"You have to register and log in to the site in order to get to the link to download the map."
I'm trying to figure out how to fix this - I'd like for everyone to be able to get to it.
It's hard to draw any conclusions without information on population density. A good way to show this is to map hue (red to blue) onto minority percentage and saturation (white to full color) onto population density. Then they highly populated areas really pop out at you while the single farmhouse fades to white.
I'll look into tweaking it to reflect density.
I may actually upload other maps later such as property values by assessed property polygon (as a reflection of household income), in addition to a pop density-only map, in order to show multiple detailed sets of polygon data.
Thanks for the critique.
Gordy expressed interest to me in learning how to use GIS.
I'm sure others are interested as well.
There's a conference next week at NCSA, monday and tuesday, that appears to be targeted at the non-technical GIS consumer community. It looks like its focused much more on spatial analysis than cartography, though, but the same methods are used in GIS to do analysis and thematic cartography.
For reference, if you think MapQuest is about the coolest thing since sliced bread, you are a non-technical GIS consumer. (No dig on MapQuest or its users, but the TIGER road shapefiles that they use from US Census leave a lot to be desired, IMHO).
http://www.news-gazette.com/news/local/2006/12/14/conference_to_focus_on_spatial_thinking.
It looks like some of the discussion will be focused on minority population spatial changes over time, something that ties in to this thread.
New map for property values vs schools loaded.
Last one I'm going to do for this.
Yeah, I saw the conference mentioned in the paper and thought I'd see about going to some of it if I could. ESRI is probably the GIS industry standard, and you can download it for free if you're part of UI. For those who can't, there are also some decent open-source GIS products, e.g. GRASS, JUMP, UDIG, and more. Actually, I'm currently working on a requirements and design document for GIS functionality in an open source product.
I've dabbled around at some of the open-source websites.
GRASS looks decent, but my big fear is that there is no product support with open-source. It's one thing to use Drupal or Linux, where there are relatively many users who can offer help, but GIS analysts, by their nature, are spread out, and far less common to run across.
My big concern is that while glitches will get fixed, there will be glitches, and no one to help me with them. There's enough glitches with ArcGIS (ESRI) as proprietary software as it is, but at least I know I can get ahold of the right people to fix it. I'm really happy with ArcGIS 9.1, now that all the patches and service packs have had time to get produced. I use 9.0 at work, and 9.1 at home, although we're about to upgrade to 9.2 at work, so I'm a little concerned about "out of the box" bugs there, since it's so new.
Yes, I know there's message boards and such for open-source, but I'd rather spend my time working on the data than the software.
On the flip side, if you (personally, wayward) are interested in it, I'd be happy to work with you on your project.
"There's a conference next week at NCSA, monday and tuesday, that appears to be targeted at the non-technical GIS consumer community. It looks like its focused much more on spatial analysis than cartography, though, but the same methods are used in GIS to do analysis and thematic cartography."
I wish I could attend.
OK, I posted small JPEGs of the maps and links to UCIMC. Here are the images in case IP wants to use them.
It's nice to see someone put this level of analysis into a post, very impressive.
I'm unsure of the what this map actually shows. As you said, the breakdown at each school is unknown, and since students may attend schools outside of their neighborhood, there's probably minimal or no correlation between demographics and school performance. Am I'm missing something?
Since no correlation can reasonably be made about the performance of a school in Districts 4 and 116 vs the non-student population immediately around each school, it's mostly just for general information purposes about the specific spatial demographic makeup of the cities, and specific locations and snapshot of academic standing of schools in the area.
I would have been shocked if there was an obvious correlation. But no, I was not making any attempt to draw a conclusion or make a point, it was just so we're all looking at the same thing, and we're all not "making up" or "guessing" our spatial and non-spatial facts about the city and the schools. Like I said at the beginning of the thread, it was in response to a seperate thread on here about the schools in the area.
(Side note: I'm interested in what Xian, JCMA, et al from that thread, have to say about the maps in relation to their discussion. It doesn't directly pertain to that thread, as far as I can tell, but it would be nice to know what they think.)
Also, this was kind of a two-part purpose. UCIMC had a similiar thread that was discussion possible locations of a new school in relation to spatial demographic makeup of the cities, so it was just as much for them as it was for IP. Thanks, wayward!
Now, the demographic makeup of each school may NOT be unknown, as I think its available on the iirc website for individual schools, and I AM able to plug in those numbers for the public schools and create tiny little pie charts at the location of each school to show relative numbers of each race at each school, over the original backdrop of minority percentage census blocks (that WOULD be useful in determining segregation in CU schools)...but honestly, I'm really kind of sick of dealing with it right now. I may do that sometime in the near future, but it's the initial data collection and data entry (especially since each school's breakdown would be on a seperate page) that's tedious and time-consuming.
Wayward:
I downloaded uDig tonight to try it out. Doesn't appear to do much more than ArcReader - am I missing something? ArcReader's already free, but has a HECK of a lot more functionality than uDig, and seems to be far more intuitive (although I'm biased in what I consider "intuitive", as a native ESRI GIS user), and has a much more professional looking and versatile layout for both viewing and print.
I looked at getting GRASS and/or JUMP to try them out first, though. I don't like the idea of a Java-based GIS like JUMP (just a prejudice, I'm not fond of Java). GRASS appears to be the most promising in terms of functionality that rivals that of an ArcInfo license, but I was really turned off when it looked like: a) I essentially have to get another program first to modify my Windows OS to accomodate what is essentially an application designed for a UNIX/Linux platform, and b) I was going to have to become a programming code expert in order to install it. Both programs looked insanely difficult to install. I've been stopped from trying out Linux multiple times for that same reason - it's made by code-saavy techies for other code-saavy techies; the unenlightened, like myself, see what it takes to install and have an "oh s**t!" moment.
Does GRASS even have a GUI toolbox, or is it all text-based tools, like the Arc/INFO line, back in the day?
They're supposedly going to improve GRASS's Windows behavior, but they may not have done this yet. (I have a MacBook for work, a Linux laptop for personal use, and a Windows desktop at work that I occasionally use.)
FWIW, here's is a list of FOSS GIS tools:
http://opensourcegis.org/
There are really a lot of them out there. At my old job, we used UDIG because it was built on top of the Eclipse framework, which we'd already decided to use for our project. It's still fairly new, and it doesn't surprise me that it's less evolved than the ESRI stuff.
BTW, I don't mind posting what I'm writing up as a draft requirements doc for a GIS-related FOSS project, but IP readers might consider it the most boring document they've ever seen.
I looked at that FOSS website - looks like there's some cool stuff on there.
There's a link on there to the Knoppix with embedded GRASS. I've tried KNOPPIX before (as a standalone OS, no GIS), and it's not too bad (a little irritated that my wireless card doesn't work on it...again, don't know the system, so I don't know how to get it to work). I do like the fact that it's a single ISO file that burns to disk, and loads directly from disk, so there's very few install issues.
That'll let me try out GRASS with my existing data on my hard drive.
If you provide more info on GIS, like your .doc, more people may get interested in GIS. I'm sure 99% of folks out their won't, but maybe one or two will! :)
He sent me this email to my UIUC address:
"Hello -- Are you the James Mortland who produced the maps related to Champaign schools?
John Gilpin, UI retiree"
I answered him yes yesterday around noonish, then asked him an hour or two later if he had a purpose in asking...still no answer.
Just kinda wierd; who is this guy? I've found a few things from googling his name, and looking it up under UIUC white pages, but not much else.
It's all good, Gilpin got back in touch with me, we figured it out.
Concerning FOSS - I've played for about an hour with the LINUX-GIS CD-ROM from that website you gave me, wayward. There's a BUNCH of FOSS tools on there, some better than others. I take back what I said about JUMP (being Java, and Java being no good). That one, at first glance, appears to be the most user friendly.
I'm going to have to spend some quality time checking out all of the programs on there.
Maybe you'll change my mind about open-source yet!
Oops, Drupal doesn't cache typed text huh :(
Anyway, great work--good data is the beginning of a more constructive discussion.
I think I'm getting the geography of the maps. The major arteries N/S are Mattis, Prospect, Neil and Cunningham, right? The blank area between the RR tracks is Spaulding?
Can you go over the "Yes/No", vocabulary again?
Even without the exact minority data at each school this is still helpful as it certainly matters whether the predominantly minority populations have convenient access to achieving schools. Choice is nice, but it doesn't eradicate transportation time issues.
I'm just eyeballing, but it's also instructive to see the greater choice more affluent, less population area have in terms of school options. This is neither surprising nor something easily addressed, as it's natural that privates would be established in areas where there is wealth to support them.
AEWS is "Acedemic Early Warning," as listed by the iirc.niu.edu website (schools in academic danger). "Spotlight," I assume, is a school that has shown some form of outstanding acheivement. "OK" is a school that had neither designator.
The field labelled "AYP?" has boolean values. "YES" indicates the school made Annual Yearly Progress, "NO" indicates the school did not, also from the iirc website.
The "Type" field is self-explanatory.
I'm pretty certain that the iirc website is basing its data directly on the NCLB, but doublecheck me. I do have other data in the attribute table for schools not shown in the map, such as school name, address, district, and what I believe the level of schooling is (k-5, jr high, high school, etc), but level of schooling was a subjective call that I made for each school, not something I actually researched, so take that for what its worth.
I actually have all schools in Champaign county listed in the attribute table, but not all of them are geocoded (many just had PO Boxes, not useful for address geocoding, and often take a lot of time per school to research actual street address), the ones that are geocoded outside of the immediate CU area were not checked for accuracy, and the ones outside of the immediate CU area have no data listed from the iirc website, since that has to be done manually (again, very time consuming), and I was only looking at the CU area.
Parts of Lincon, 130 and Staley Rd are in red as well (N/S), and the major E/W road is Springfield in the West, and University in the East. 150 is also in there on the NW corner. Interstates should be self-explanatory, and you should be able to ID some of the water features like Crystal Lake and the Boneyard.
Tip for next time: I need to label major roads.
Sorry, I'm not sure what area you're referring to as "Spaulding." If it's blank in the minority percentage map, no one lives (well, liveD - in 2000) in that block. If it's blank in either of the other two maps, either no one lives there (all) , or it's in the lowest density category (density only), or something was screwy in my definition query (property values only).
Yeah, I wasn't sure what that "Yes" "No" column was for, thanks for clearing it up.
Here are the standards for "Spotlight" (this is old, but should still apply with the years changed):
The 27 Illinois Spotlight Schools meet the following criteria:
Sorry, I shouldn't have used shorthand--Spaulding=Spaulding Park, so since no one lives in the park (officially anyway) I figured that would account for that blank area.
Note how the RR tracks influence the racial geography, IIRC correct, the river serves a similar purpose in Rockford. Interesting how we unintentionally use geography to segregate ethnic minority communities.