A Bottom-up Data Culture?

[Graham Lally]

Hi list, long time no see. Hope you're all having a great Christmas.

Been having an interesting little conversation today with James Cattell
(@jacattell) and Si Whitehouse (@siwhitehouse) on Twitter about what
makes open data attractive to users.

[TLDR - should/can we make open data more about starting with users'
data, rather than seeing them as 'consumers'?]

The original context was a tweet about a general lack of "armchair
auditors", which have generally failed to appear as the phrase might
suggest. (Although it's great to see more journalists realising how
important data is.)

So far, Open Data has concentrated on helping people with "data" to make
it more open - the 5 stars, the notion of transparency, data silos, etc,
all assume that data is held by organisations and "we, the people" want
to access it.

But the move towards a world of data as an infrastructure is progressing
slowly. And maybe with good reason - can we really expect a political
data infrastructure if there is no broader *culture* of data? And can we
expect such a culture to appear from a central-outwards, broadcast model
of how data gets generated and used?

Should there be more focus on helping people to engage with their *own*
data - or, at least, data relevant to them? By "people", I mean anyone
who wouldn't normally consider themselves to use "data" (open or
closed), but that deals with (structured?) information on an every day
basis?

For example, I'd love to see a simple checklist of questions that help
people/groups just *identify* the data they have, what might be useful
to others, what could be derived from it, and possible ways to open it
up. Simple questions.

James is also up for the idea if open data surgeries, in the same style
as social media surgeries. Maybe just a list of "data mentors" for
advising on data issues might be useful?

There is a tendency, imho, to descend into a personal data fetish view
of the world (in which everyone is encouraged to record every detail
about their life and share it). I'd like to avoid such a heavily
technical perspective, and aim for:

a) more of an awareness about the potential of the information/data we
already use - either that generated by users, or what they actually use

b) creating simple solutions that draw on this potential, developed in
conjunction with the users

Is this crazy talk? Maybe we already have armchair auditors, just they
work for existing organisations, using existing tools. Maybe a data
culture is coming on faster than I think it is and I'm just impatient.

Either way, I feel like making 2013 the year of "data culture"...

Happy new year,
Graham

[Giuseppe Sollazzo @puntofisso]

Hi Graham,

On Thursday, 27 December 2012 23:16:47 UTC+1, Graham wrote:

[TLDR - should/can we make open data more about starting with users'
data, rather than seeing them as 'consumers'?]

my general feeling is: we should and we should not. Indeed, the two processes should be part of the same team advocating for Open Data within an organization.

 

So far, Open Data has concentrated on helping people with "data" to make
it more open - the 5 stars, the notion of transparency, data silos, etc,
all assume that data is held by organisations and "we, the people" want
to access it.

But the move towards a world of data as an infrastructure is progressing
slowly. And maybe with good reason - can we really expect a political
data infrastructure if there is no broader *culture* of data? And can we
expect such a culture to appear from a central-outwards, broadcast model
of how data gets generated and used?

I think that a first phase of "we, the people, want to access the data" is vital and required in order to establish a proper culture of openness. Without that, it's difficult to think of a sustainable economy of Open Data. Also, as organizations and institutions it's already rather tricky to find resources and collaborators in-house to drive advocacy projects for Open Data. I keep telling everyone how frustrating is to deal with FOI officers whose job seems to be more to prevent Openness than to facilitate it.

My work is in academia. In academia, most of the groundbreaking work for Open Data in many institutions has been initiated by people who were passionate about that (even in Southampton). There's never been an institutional drive towards Open Data, rather a passive acceptance of it if not an explicit hostility towards Open Data projects (often quoting "security" or "legal" issues).

In such an environment, it's clear that the positive drives for Open Data have been two:

- Open Data champions, internal to the organization

- "we, the people" pushing for such data to be opened.

 

Should there be more focus on helping people to engage with their *own*
data - or, at least, data relevant to them? By "people", I mean anyone
who wouldn't normally consider themselves to use "data" (open or
closed), but that deals with (structured?) information on an every day
basis?

This is a great idea, but I struggle to see how that could be run by organizations like mine, or even in local government. 

What would be needed is a complete reform of what FOI officers are today, and it's hard to see that happening. In my case I talk often with people doing research, for example, willing to advertise their job. In many cases, these people own data which could be shared externally. What we need is a way to present this people with reasons and motivation to work with these data as Open Data. And for this, your next point is a good starting point.

 

For example, I'd love to see a simple checklist of questions that help
people/groups just *identify* the data they have, what might be useful
to others, what could be derived from it, and possible ways to open it
up. Simple questions.

James is also up for the idea if open data surgeries, in the same style
as social media surgeries. Maybe just a list of "data mentors" for
advising on data issues might be useful?

Data mentors, data surgeries, opening data checklists - all help. We need ways to show people owning data the benefit in opening these - and to drive the need for openness by the external players.

One question: what if there is data available but no one interested in it? Do we still have a "duty" to try and open it if relatively economical to do so?

 

There is a tendency, imho, to descend into a personal data fetish view
of the world (in which everyone is encouraged to record every detail
about their life and share it). I'd like to avoid such a heavily
technical perspective, and aim for:

 a) more of an awareness about the potential of the information/data we
already use - either that generated by users, or what they actually use

 b) creating simple solutions that draw on this potential, developed in
conjunction with the users

Is this crazy talk? Maybe we already have armchair auditors, just they
work for existing organisations, using existing tools. Maybe a data
culture is coming on faster than I think it is and I'm just impatient.

Not crazy talk, at all.

But we need to distinguish between personal data and non-personal data (even considering that non-personal data can lead to identifying people). Opening Data is a process that requires data to be accessible. Then we can discuss policies by which we define who's entitled to access what. It's not just about the technicalities.

 

Either way, I feel like making 2013 the year of "data culture"...

Happy new year,
Graham

I hope so :)

Happy new year,

Giuseppe @puntofisso 

[Tim Davies]

Hello Graham and all,

Happy New Year. Thanks to Graham for sparking some conversations on this list again. 

I'm definitely in favour of much more focus on building a (critical) culture of engagement with data - looking a up-skilling people as well as the focus on tool and platform-building and intermediaries that has tended to predominate in the dialogue so far. 

It would be great to explore ways to catalyse this more: I think a lot of the jigsaw pieces might be around (I'm talking with Nick Booth tomorrow about a project with an open data surgeries angle; Nominet Trust and Big Lottery fund are looking at some sort of action learning process with Third Sector around open data this year; the open data cook book and Open Data Engagement stars are a bit dormant but good traction; the experience of DataBridge and other projects provide great foundations to build on), but we've not yet got a good way to put them together or drive them forward...

Are we thinking a session at UKGovCamp to catalyse some action? Or worth trying to convene a wider conversation through other routes?

All the best

Tim

--


http://www.timdavies.org.uk
07834 856 303.
@timdavies

Co-director of Practical Participation: http://www.practicalparticipation.co.uk
--------------------------
Practical Participation Ltd is a registered company in England and Wales - #5381958.

[Graham Lally]

Thanks for the replies Tim and Giuseppe, and hope everyone had a great
new year.

Reading through both mails, it really feels like the challenge for 2013
is to help Open Data to escape from a "one size fits all" approach at
various levels (which is probably driven by a push from the center,
politically, financially and publicity-ally).

That's not to say current efforts are "wrong", but rather that current
efforts need a grander perspective than simply saying "an institution
should provide data in an open format for everyone". The nuances of data
usage are more subtle than what we've been dealing with so far.

I'm gearing more towards a data market approach, personally, in that the
supply and demand of open data is often very subjective, and the
appropriateness of the technology being used depends on both the creator
and the consumer, *and* (or more precisely) what they want to do with it.

(Similarly, scraping the Twitter API for all your tweets can work
technically, but there are many reasons why Twitter are making archives
available through CSVs...)

A while ago I started on a kind of "map" for open data usage, but it got
complicated quickly. I might dig it out and try to simplify it soon
though, as it was quite useful to try to identify who was likely to want
data from whom and how, and might help with this idea of a "broader
conversation", as well as with helping what *kind* of Open Data is most
valuable to different parties.

Definitely up for some kind of development at UKGovCamp (and on the
list, of course) - the first question, as always, is "where do we
start?", and whether that should be a "theoretical" start (frameworks,
etc?) or a "practical" one (develop a tool to develop ideas).

From the "supply-vs-demand = success!" perspective, I'd love to map out
somehow the overlap between data available and data wanted, to try to
make sense of what makes an open data ecosystem actually valuable.
Hopefully this also starts to answer Giuseppe's question of "duty to be
open"...

Projects like DataBridge and Making a Difference with Data might be good
starting points for this. As Tim notes, tying it all together is still
an open action point.

Any other ideas?

Best,
Graham

[Graham]

This sounds like an interesting exercise too:

http://schoolofdata.org/2013/01/03/join-the-first-online-data-expedition-flatlan

d-a-romance-of-heavy-dimensions/

I like the way it 'sells' data as exploration and storytelling, rather than a

duty. More about the Data Expeditions here:

http://schoolofdata.org/data-expeditions/

- Graham

[Giuseppe Sollazzo]

Hi Graham,

On Thu, Jan 3, 2013 at 9:16 PM, Graham Lally <exm...@gmail.com> wrote:

Reading through both mails, it really feels like the challenge for 2013
is to help Open Data to escape from a "one size fits all" approach at
various levels (which is probably driven by a push from the center,
politically, financially and publicity-ally).

That's not to say current efforts are "wrong", but rather that current
efforts need a grander perspective than simply saying "an institution
should provide data in an open format for everyone". The nuances of data
usage are more subtle than what we've been dealing with so far.

I agree on both counts.

The problem with the "one size fits all" descends from the abstractness of the FOI legislation, though: right to data access is granted - *how* to do this is left to each institution without providing guidance. Surely it would be problematic should FOIA include technical requirements, but the total lack of it doesn't help either. See also what I write later about FOIA Section 21.

This is made evident by your next observation.

 

I'm gearing more towards a data market approach, personally, in that the
supply and demand of open data is often very subjective, and the
appropriateness of the technology being used depends on both the creator
and the consumer, *and* (or more precisely) what they want to do with it.

What happens if different consumers require different "presentations" of the same data? 

What should be the duty of the institution to accommodate such different, but similar, requests?

 

A while ago I started on a kind of "map" for open data usage, but it got
complicated quickly. I might dig it out and try to simplify it soon
though, as it was quite useful to try to identify who was likely to want
data from whom and how, and might help with this idea of a "broader
conversation", as well as with helping what *kind* of Open Data is most
valuable to different parties.

This is something that could form part for a discussion. "Who is likely to want data from whom and how".

I'm not a legal expert, but I always wonder how the discourse on Open Data should consider the provisions of the FOIA Section 21: http://foiwiki.com/foiwiki/index.php/FOIA_Section_21_Exemption - this seems to suggest that data that are already available are exempt by FOIA requirement. 

This might open an interesting discussion: are we going to see Open Data releases being used as an excuse to turn down FOI requests?

 

Definitely up for some kind of development at UKGovCamp (and on the
list, of course) - the first question, as always, is "where do we
start?", and whether that should be a "theoretical" start (frameworks,
etc?) or a "practical" one (develop a tool to develop ideas).

From the "supply-vs-demand = success!" perspective, I'd love to map out
somehow the overlap between data available and data wanted, to try to
make sense of what makes an open data ecosystem actually valuable.
Hopefully this also starts to answer Giuseppe's question of "duty to be
open"...

There seems to be more than enough material for a discussion, and I'd be happy to hear what other Open Data champions think about, their experience in their own institutions, etc...

G

--
Giuseppe Sollazzo
- http://www.puntofisso.net
- http://twitter.com/puntofisso

My newsletter is available at: http://tinyletter.com/puntofisso

"Gorau chwarae, cyd chwarae'"

[Graham Lally]

Quick follow-up...

On 03/01/2013 21:16, Graham Lally wrote:
> A while ago I started on a kind of "map" for open data usage, but it got
> complicated quickly. I might dig it out and try to simplify it soon
> though, as it was quite useful to try to identify who was likely to want
> data from whom and how, and might help with this idea of a "broader
> conversation", as well as with helping what *kind* of Open Data is most
> valuable to different parties.

Here's my original 'map' of data suppliers and users, put together after
last year's UKGovCamp session:

https://docs.google.com/open?id=0B3MuKQGBY6LhWGNSNGgxWGprM0U

From trying to identify the various actors, it then adds in the various
functions that different groups are most likely to want to perform on
data. Mapping *that* is either really simple, or really complicated,
depending on how much you want to cross-reference it all...

No idea if this is useful or not.

Graham

[Tim Davies]

On Sat, Jan 5, 2013 at 12:20 PM, Giuseppe Sollazzo <punto...@gmail.com> wrote:

Hi Graham,

I agree on both counts.

The problem with the "one size fits all" descends from the abstractness of the FOI legislation, though: right to data access is granted - *how* to do this is left to each institution without providing guidance. Surely it would be problematic should FOIA include technical requirements, but the total lack of it doesn't help either. See also what I write later about FOIA Section 21.

For any who have not yet seen it - government are currently consulting on the Code of Practice to accompany the Right to Data in the updated FOI Act (ending 10th Jan)

The current Code is generally I think seen to be not very good by many advocates of open data, and there might be some real scope for feedback to influence it's development:

http://data.gov.uk/consultation 

The Code falls short of richer guidance, but its eventual launch (later this year?) might provide a good moment for some community developed guidance, or for one an organisation like LGA etc to be encouraged to share richer guidance...

All the best

Tim

 

[Graham]

Hi Giuseppe, all,

Quick note to say that James has brought up the discussion over on the
UKGovCamp 2013 list, for ideas for anyone attending. I've posted a reply
there about maybe coming up with a series of questions for people to start
working with/opening up data in a demand-led way:

https://groups.google.com/d/msg/ukgc13/0bjJIfiBKNI/ipjmseXj1KMJ

And included here for discussion:

Thanks for re-posting my ODE post, James. Got a brief moment to pick up on
one small aspect I think I'd like to look into at UKGC, which ties into...


On Thursday, January 3, 2013 8:16:44 AM UTC, Mark Braggins wrote:
Focus on demand - if orgs are releasing data and there are indications
that it's being used, do more of that

Demand is key, imho - or in other words, how do we start with (open)
*questions* to make the (open) answers/data have value?

I'd probably set out some base assumptions to stop the fairly vague topic
getting more vague (Open Data often ends up trying to solve too many
things in too small a session):

1. Technology can be generic, but discussion needs to start focused -
always work within a context, ie. tackle a real-world problem with
real-world actors and data. (Expand outwards if needed.)

2. The *demonstrable* value of open data lies in answering questions, not
in any grand principals of openness. All discussion needs to be framed
within a "so what?" context too.

From this, I'd be interested in working out some very practical "open data
starters" for anyone wondering how it all applies to them. For example, a
set of questions such as:

1. What do you want to do?
2. What data do to need to do it?
3. What data do you have?
4. What data do you need?
5. ...

etc... This could break down into more focused questions, and cover other
related topics such as skills needed, etc.

The overall aim would, as set out above, hopefully provide a more
practical guide to why open data is useful (for *anyone* interested in
data - citizens, schools, etc as well as local gov, researchers, etc), in
a format that might be both useful and easy to explain to people (as well
as to update).

You may notice that there's not much there about "openness". This is
because I see the openness as something which emerges from people and
organisations working together in an open fashion, rather than aiming for
"generic openness" from the start. Start to ask "where can you get data
from?" on a large enough scale, and see if openness appears.

My 2p anyway :)

- Graham