Good point Tim. We effectively do lots of this kind of thing at workshops and events run by us and others.

E.g. http://blog.okfn.org/2011/08/09/data-driven-journalism-workshop-on-eu-spending-tools-techniques-utrecht-8th-9th-september/

And numerous crash courses and clinics in open data licensing, using CKAN, etc.

I guess we could formalise some of these into short courses. E.g. in association with P2PU. And start to get more systematic about engaging with new stakeholders - e.g. in universities, public bodies, media organisations, interest groups, etc?

Best way to learn using open data is to have a package that is easy to install & hack it. There are some downloads offered currently but the hurdle to hack it is still high. Some nice things to have would be maybe a image of a running instance that can be locally started to begin hacking, have a cloud instance that can be deployed (on amazon,...) and tried out... Having youtube videos to show usage will help offcourse together with P2PU content as well..

Sure lot of volunteers will be needed for this but having real projects (ex: http://apps4deutschland.de/) where the software stack can be tried out may be more practical way to approact this

My2Cents Srini

i think this is a great idea and has a lot of potential. I like the idea of collocating a bunch of resources, tutorials, etc together for the purpose of educating people how to use, evaluate, understand, etc open data.

I do have a few questions. I think the most important question is who would something like this be geared towards? Who is the primary audience? Would this be geared towards total beginners (zero knowledge or limited skills) or would this be geared towards people that already have some basic knowledge & skills (programming, stats, etc)? Would this curriculum be primarily designed as a sort of tech "cookbook" for how to use data or would it also cover basic issues like information literacy, statistical literacy, data literacy?

Some more comments based on my thoughts from IRC discussion

I think the manual is a good first step and then from the manual you can start to build a more robust curriculum around it.

For open data we are not just teaching people how to use open data we also need to teach them how to find, understand, critically evaluate data (think of it as data literacy similar to info/stats literacy). We are teaching them how to be "data literate" so that can be better consumers and users of open data.

So i think you have two primary types in terms of audiences 1) those that are some what familiar with data and could do more if they just learned a few more skills; 2) those that have little or no experience with data (which is probably the overwhelming majority of the pop). The former is easier to designed for but may not have the most bang for the buck (ie preaching to the choir; most tech people can possibly find other more appropriate resources elsewhere). The latter is much harder since its working with people that have poor info literacy, stats literacy or little tech skills. However the latter is where you could envision the biggest gains for open data. You are teaching people who previously didn't understand data to understand and appreciate data. If more people understood and appreciated data they would be more like to advocate for open data (converting new followers). This also helps ensures that open data is truly "democratizing" rather than just "liberating" and that open data has positive outcomes and does not work to further disenfranchise people. Educating non-technical citizens helps broadening the community and helps to guarantee more meaningful access for everyone with the added benefit that it possibly helps create more demand and more advocates for open data movement.

The goal shouldn't be to get everyone to be "users" of open data since not everyone can or will be active users of "raw" open data (think 90/9/1 principle; 90% will be consumers of derivatives of open data; 9% will be advanced consumers and do some "advanced" things with data but not much; 1% will be users of raw open data to make things digestible and useable for the other 90%).

As for curriculum, I could see curriculum topics being logically organized and structured around real world activities (as in i want to do...)

How to find talk about open data vs closed data how to find data where to find open data how to request advocate for open data How to use <- big category; lots of ways to use data effectively a tech cookbook provide list of tools provide ideas for using data (maps, visualization, statistical analysis, programming applications, etc) provide real world examples; real world exercises How to evaluate take material from information literacy, statistical literacy curriculum How to promote/advocate advocacy materials for promoting open data how to get your organization to open data; what to say; talking points

A good road map for creating curriculum would be...

Steps for the construction of curriculum Define audience (audiences) <- often ignored Determine the need <- often ignored Define goals and objectives of curriculum to address needs/deficiencies Selecting determining subject content Structure and prioritize topics

I could see the curriculum basically structured around specific topics with real world examples; after each topic and example; there should be an activity or a real world exercise that people can try/practice; learn by doing; for example take this data and what you have learned from this lesson make a map or an application; or look at this data and evaluate it for potential problems and errors; look at this information based on data, identify what is wrong with it.

After curriculum is finished being constructed you can then release it under cc licenses so that it is not only free and publicly available but also so that people can build off of it and improve it. Goal should be something under the guise of an open courseware approach (i.e. create curriculum that other people can use to teach). I think if the curriculum is open, I'm sure you would have a much easier time trying to promote teaching this material in formal and informal settings since one of the hardest parts for teaching is crafting the curriculum.

After the curriculum is constructed you can take several approaches to advocating the use of the curriculum. You have two channels: informal learning environments and formal learning environments.

1) informal; make material available online as both a resource and as a series of tutorials; make videos that assist people step by step ala khan academy; designed for informal learning; people can go to websites learn at their own pace; "guerrilla learning" or life long learners.

2) formal - partner with existing organizations and institutions to help advocate for curriculum and to help inject curriculum into existing courses.

K-12 - Work with public schools to encourage the use of curriculum when teaching students about data; frame this as another tool for students and teachers to support their existing work;

University - Work with universities, professors, professional organizations (ACM, etc) to encourage the use of curriculum in classes (policy, stats, data viz classes, etc)

Adults & life long learners - Public libraries (ubiquitous; existing infrastructure w/ teachers, tech, etc); Community centers; Government sponsored training events; Community sponsored workshops host classes not hack-a-thons; don't do, teach people how to do;

Has anyone considered if making all content public domain or cc0 is a better option than cc..? http://tunes.org/legalese/bugroff.html

{I tried to leave this as a comment on justgrimes post but the comment button didn't work (I have flashblock enabled}}

I think the best idea for teaching is a clear step by step guide of 'Howto' carry out a task for each step. Ideally leading anyone from start to finish with minimal further input and with total success.

It may be best not to get too drawn into how to organise a persons learning. If the content is first class and freely available then that could be of more importance.

Perhaps consider the free school (http://en.wikipedia.org/wiki/Sudbury_school) ideas vs. anything formal such as a curriculum. If the content is easy to follow (visually and logically) then how people are guided to learn is possibly of less importance, as 'search' will be the guide.

If possible link to other useful content rather than rewrite it (e.g., http://wiki.debian.org/DebianIntroduction).

[btw the last step is where things get sticky wrt licenses. One has to ask is my license compliant with theirs, do their links have licenses that are compliant, does it matter? etc.]

I'd love to be able to hack on open data projects focused on ICT4D on my ubuntu localhost maybe within a Pinax/Geodjango environment. http://www.paolocorti.net/2009/10/10/the-pinax-tutorial-1-installing-pinax-and-making-basic-customisation/

So a github repository that can be installed locally or to a live environment.

Once the environment is established I'd like to follow a curriculum based on 'specific topics with real world examples'. Maybe we could establish KPI's based on Millennium Development Goals.

http://www.unesco.org/water/wwap/facts_figures/mdgs.shtml

Yes a structured 'How to' manual would really help. As an artist interested in making work using Open Data (see http://ideas.okfn.org/ideas/46/landscape-portrait) who has a rudimentary understanding, there is a really steep curve in order to move from thinking to making and using, anything that can alleviate that would be great, I am happy to act as a guinea pig. Kevin