The Atlanta Journal Constitution reported that Brandon Elementary scored top in the District of Atlanta Public Schools and third in the state on the standardized tests in Georgia. William Fragakis and I (posted by Daniel) converted all the old, nonfunctioning PCs into Linux thin clients running K12LTSP, got more than 100 donated PCs from businesses, and bought more than 60 new NTAVO thin clients at $99 each. The ended up more than tripling the number of working PCs in each classroom.
Scores were up across the board, and math scores, in particular were up sharply this year. All of the teachers have told us how much having all the Linux PCs have improved things, but this is the first hard evidence. Not that I'm a complete fan of all the testing that goes on here lately, but test results are paid attention to by decision makers, many of the ones we want to convince of the benefits of open source software and thin client architectures.
Thanks to the group for all the support! Daniel
Thursday, June 22, 2006
Sharp increases in testing results for a school after switching to Linux
http://www.ajc.com/metro/content/metro/stories/0621metcrctbestchart.html
Subscribe to:
Post Comments (Atom)
Posts
1 comment:
Could you tell us in more detailed way that how "Linux+thin client" can improve math score?
Here's what we found: Moving to Linux and thin client architecture allowed us to cost effectively triple the number of working PCs in each class. This, coupled with more reliable web access (via a local cable modem) allowed the teachers to *regularly integrate* the use of the PC into daily activities, which they were not doing before. In addition to the built-in OSS math applications, the teachers used web sites such as First-In-Math on a daily basis for all students. The combination of
daily, structured, individual time on the PCs doing reading and math
activities, coupled with the use of PC math activities when students finished their work early (and previously had to be kept out of trouble from distracting others) increased the time on task due to 1) the availability of ubiquitous applications (either web based, or OSS loaded on all servers) and 2) enough working PCs in each class so that all students could use them daily for at least an hour, and also in between other activities. This produced a massive increase in the time spent doing math activities for all students. The kids thought they were playing games or doing a web competition, but really they were
practicing math skills in a thoroughly enjoyable manner to them.
Further, improved productivity using the PCs led to more time available to practice math: the conversion of creative writing activities from pencil and paper to PC based (using Abiword) more than halved the time it took to produce a final product, again, leaving more time for math activities in between.
Another effect we saw was that at the end of the year, the 5th grade teachers began an inquiry-based project on the ocean in the last two
weeks of school. They really just turned the students loose on the PCs
with some guidance on the type of info to find, and the students found it themselves and produced bulletin board displays, presentations (teaching themselves Impress in the process), etc. The teachers stated that they would have never tried to begin such a project in the final weeks of school before, it would have taken 2 months the old way without enough PCs, and used to be the last two weeks were sort of a ramping down period. The principal stated that in all her years of teaching/administrating, she had never seen the last two weeks of school be so productive and filled with intensive learning.
So, I'd say in a nutshell that the OSS and web based software apps made learning math fun for the students, and having enough PCs, which was only possible using OSS and the thin client architecture, meant that all kids had plenty of access to PCs and could seamlessly use them throughout the day. Results were increased productivity, increased time
spent doing math, and extension of learning right up to the final days
of school.
Daniel
- - - -
I'm curious which built-in FLOSS math apps were these & which grade levels?
For middle-schoolers I've found three upper-level non-FLOSS math apps that are portable across Windows & Linux, namely MathAid's java/javascript algebra & trig lessons & Maple, but I was wondering what if your teachers found any FLOSS alternatives that were helpful for middle school students.
Our highest grade is 5th, so we used KBruch, Kpercentage, TuxMath, Gperiodic, and Kalzium for the older kids and for the younger grades, Gcompris and Childsplay were big hits.
Since the district is now planning to do 6 more schools, including middle schools, we should be able to begin exploring what middle school kids groove on. I know there is an OSS Maple alternative as well as for
Matlab, one of my favorites that I think would work for middle schoolers. Here's a fairly complete list of math OSS apps from
http://www.linuxrsp.ru/win-lin-soft/table-eng.html#9
A couple of the Linux apps are proprietary, but are noted as such. SEUL app index also has plenty of algebra/geometry apps if memory serves.
Daniel
Math system in MathCad style: Gap.
Math system in Matlab style:
1) Matlab. [FTP]
2) Octave. (+ Gnuplot)
3) Scilab.
4) R.
5) Yorick.
6) rlab.
7) Yacas.
8) Euler.
9) Freemat
Math system in Mathematica style
1) Mathematica. [Prop]
2) Maxima.
3) MuPad.
4) NumExp.
5) Mathomatic.
Math system in Maple style
1) Maple. [Prop]
2) Maxima.
3) MuPad.
Equation / math editor
1) OpenOffice Math.
2) MathMLed.
3) Kformula (Koffice).
4) LyX.
5) Texmacs.
Engineering
1) Varkon.
2) Linuxcad. [Prop, ~100$]
3) Varicad. [Prop]
4) Cycas. [Prop]
5) Tomcad.
6) Thancad.
7) Fandango (alpha-version).
8) Lignumcad.
9) Giram.
10) Jcad.
11) QSCad.
12) FreeEngineer.
13) Ocadis.
14) PythonCAD.
15) OpenCascade.
CAD/CAM/CAE: Qcad.
Try this one, it shows Brandon as the top Atlanta school, but doesn't
show the summary chart in the paper where we're listed third. Daniel
http://www.ajc.com/metro/content/metro/stories/0621metcrctbestchart.html
Post a Comment