Monday, April 28, 2008 at 5:36 PM
Last week, we excitedly announced that we'd accepted over 1100 students into Google Summer of Code™ 2008. Now that we've had some time to get everything in order for this new crop of students, we're happy to bring you some of the news you've been asking for about this year's participants.
Even though this is our fourth summer together, it seems the statistics folks are most interested in remain the same: who are the participants, what are they working on, who were the applicants and what places do all these folks call home? You can find answers to the first two questions on the program website, and you may remember that we have students and mentors from 98 countries this year. With so many data points to choose from, we figured that we'd start this week off with a look at two Top 10's for 2008: number of applicants by country and number of accepted students per country. Not surprisingly, the distribution for both sets looks remarkably similar:
.png)
Now that we've whet your appetite with a bit of regional data, you'll be pleased to know that we've got even more for you. We'll be producing the usual KML file showcasing our 2008 'graduates' at the end of the program, but thanks to our stalwart intern and organization administrator for the Wine project, Maarten Lankhorst, you can already check out maps linking students to mentors from each individual project's page on the Summer of Code site. Still not enough data for you geo-wise? Check out the our hall of fame maps from 2007 (KML), 2006 (KML) and 2005.
We're curious to know what information you'd like to see us showcase about Summer of Code. Why not post a comment and let us know what questions you have?
Ed. note: Updated post with corrected graph.
15 comments:
Hey Leslie,
I'd like the see the distribution of repeat-students that are participating broken down by years of experience in the program (0-3 years).
Thanks,
Kris Katterjohn (Nmap Security Scanner)
are those graphs actually different?
Dear LH & SoC Team
The graphs clearly shows The increased no of accepted student applications from India for summer of code. As per a rough calculation Last year it was nearly 60- Projects and This year it is 70 + . But while looking at the geographic base of Mentoring organisations there is no mentoring organisation from India other than Gnowledge project(participating as a part of GNU project) . I think you also need to look on this issue a bit seriously nexttime . An India specific GSoC students and mentors list is compiling in http://fci.wikia.com/wiki/SoC/2008
Hi Leslie, thanks for breaking down the applications by country. I don't expect you do crunch all these numbers just because I suggest them, but there are many other breakdowns that I think would be very interesting:
* Average Student Age by Mentoring Organization. The average age of applicants and the organization to which they applied, and likewise for the average age of successful applicants. This would help organizers understand if some groups are biasing their selection process towards older students.
* Top 10 universities. Which universities have the largest absolute number of students participating in summer of code, and among those with at least N students participating, which universities had the greatest acceptance rate?
* Repeat-student distributions. Broken down by project, or by age, or other factors.
* The same student/country data, but imported into Gapminder / Trendalyzer over each of the last 4 years, so that the data can be visualized along with changes in GDP, internet connectivity, literacy, and other socio-economic factors.
Exactly, except the colors is there any difference between above two graphs ??
Anyway, I think numbers(it can be in percentage) along side the sections of pie graph would be helpful.
Then, like you can say to people that *this* percentage of students participated from *this* country in gsoc.
It would be nice to see number of students from each country. Thanks for info about TOP 10 countries, anyway.
@jaltmna: I've compared differences in GIMP. These graphs are exactly the same. o_O
BTW: it would be great to see some numbers on these graphs.
Thanks for TOP 10 countries anyway :)
Noticed the map connecting the student and the mentor. Way too cool. Kudos to you, Maarten Lankhorst.
LH, great to see stats. I am pleasantly surprised to see the number of SoC'ers from India.
Few other ideas for the charts:
- Chart for all the applicants, not just top 10 would be of interest as well.
- Comparison charts from previous year SoC'ers.
Thank you,
Senthil
SoC'er - 2007 and 2008, flying the Indian flag.
Hey Leslie, all.
>are those graphs actually different?
I used gimp to compare: they're not different. But scale is too low to see the difference: 1° is used for 3 students.
>I'd like the see the distribution of >repeat-students that are participating >broken down by years of experience in the >program (0-3 years).
+1. Also interesting to see how many participants accepted this year were not accepted last (and vice versa).
Perhaps percentage of female vs male?
I think it would be really interesting to see applicants / accepted students by continent :)
I find it surprising that the two pie charts look identical, the charts suggest the percentage of students from each country who applied to the program appears to be equal to the percentage of students from each country who were accepted to the program.
I'd like a dataset containing an id for each individual, their country, and organization applied to for both applicants and accepted students. The GSoC community is likely to have more time than the GSoC team at Google to tease interesting results from the data.
Thanks to those who pointed out that the graphs were the same - now fixed.
Thanks also to everyone for their suggestions about what stats they would like to see in the future; we'll be featuring updates addressing some of these requests in the coming weeks.
This is really interesting. Sri Lanka has reach 9th place in both lists; well done.
Congrates on india. Well done.
Post a Comment