open-discussion
open-discussion > Suggestion/plea: stackoverflow-like QA site
Mar 11, 2011 03:03 AM | Isaiah Norton
Suggestion/plea: stackoverflow-like QA site
Does something like stackoverflow exist for neuroimaging? The only
thing I've seen is this recent proposal on stackexchange: http://area51.stackexchange.com/proposal...
...but it's still in staging. Are there any others? Maybe the above site is the right way to go. But, if NITRC is interested in hosting such a site there is a good open-source implementation: http://www.osqa.net/
(the rest below is general thoughts and opinion...)
The forum model is great for some types of communication - exhibit A: today's continuing discussion about the new AAAS Editorial policy. Such conversational topics are perfect for a threading forum. However, the forum and project-specific mailing lists may be suboptimal for well-defined question/answers given the availability of a potentially better option.
The stackoverflow model has become tremendously popular and is applicable beyond programming for many reasons. To give a few science-related examples:
http://qa.nmrwiki.org/ (NMR community and QA site)
http://math.stackexchange.com and http://mathoverflow.net for general and research math respectively.
http://physics.stackexchange.com/
http://metaoptimize.com/qa/ and http://stats.stackexchange.com/ for machine learning and (somewhat) more general stats.
...but it's still in staging. Are there any others? Maybe the above site is the right way to go. But, if NITRC is interested in hosting such a site there is a good open-source implementation: http://www.osqa.net/
(the rest below is general thoughts and opinion...)
The forum model is great for some types of communication - exhibit A: today's continuing discussion about the new AAAS Editorial policy. Such conversational topics are perfect for a threading forum. However, the forum and project-specific mailing lists may be suboptimal for well-defined question/answers given the availability of a potentially better option.
The stackoverflow model has become tremendously popular and is applicable beyond programming for many reasons. To give a few science-related examples:
http://qa.nmrwiki.org/ (NMR community and QA site)
http://math.stackexchange.com and http://mathoverflow.net for general and research math respectively.
http://physics.stackexchange.com/
http://metaoptimize.com/qa/ and http://stats.stackexchange.com/ for machine learning and (somewhat) more general stats.
One evidence for the potential utility of such sites
is the presence of some very high-profile contributors on
mathoverflow (eg Terence Tao).For those who may not be
familiar with the model, I think the best analogy is "an
interactive FAQ". A few key features contribute to the
self-reinforcing usefulness of these sites:
1. users get reputation points for answers.... It's behavioral economics in action, and it works.
2. community moderation improves signal/noise, with privileges based on reputation.
3. questions and answers are re-editable and have specific comment sections. This compares favorably to the annoying correct/refine/update traffic for some questions on a mailing list.
4. convertability: answers can easily be turned into wiki pages when appropriate to provide evolving reference sources for fluid answers.
5. questions and answers are tagged by subject area (ie: programming language or branch of math in the above examples)
6. language agnosticism.
I think 6 is especially critical. The neuroimaging corollary - software and toolkit neutrality - would ideally attract answers from participants in many different projects. This could help raise awareness of alternative tools and workflows in a way that project-specific mailing lists cannot. NITRC has been a great development in this direction, and a neuroimaging-related QA site could push further.
1. users get reputation points for answers.... It's behavioral economics in action, and it works.
2. community moderation improves signal/noise, with privileges based on reputation.
3. questions and answers are re-editable and have specific comment sections. This compares favorably to the annoying correct/refine/update traffic for some questions on a mailing list.
4. convertability: answers can easily be turned into wiki pages when appropriate to provide evolving reference sources for fluid answers.
5. questions and answers are tagged by subject area (ie: programming language or branch of math in the above examples)
6. language agnosticism.
I think 6 is especially critical. The neuroimaging corollary - software and toolkit neutrality - would ideally attract answers from participants in many different projects. This could help raise awareness of alternative tools and workflows in a way that project-specific mailing lists cannot. NITRC has been a great development in this direction, and a neuroimaging-related QA site could push further.
Threaded View
| Title | Author | Date |
|---|---|---|
| Isaiah Norton | Mar 11, 2011 | |
| Daniel Kimberg | Mar 11, 2011 | |